ALBERT

Description: Application of machine and deep learning methods in drug discovery and cancer research has gained a considerable amount of attention in the past years. As the field grows, it becomes crucial to systematically evaluate the performance of novel computational solutions in relation to established techniques. To this end, we compare rule-based and data-driven molecular representations in prediction of drug combination sensitivity and drug synergy scores using standardized results of 14 high-throughput screening studies, comprising 64 200 unique combinations of 4153 molecules tested in 112 cancer cell lines. We evaluate the clustering performance of molecular representations and quantify their similarity by adapting the Centered Kernel Alignment metric. Our work demonstrates that to identify an optimal molecular representation type, it is necessary to supplement quantitative benchmark results with qualitative considerations, such as model interpretability and robustness, which may vary between and throughout preclinical drug development projects.

Description: Circular RNAs (circRNAs) are widely expressed in highly diverged eukaryotes. Although circRNAs have been known for many years, their function remains unclear. Interaction with RNA-binding protein (RBP) to influence post-transcriptional regulation is considered to be an important pathway for circRNA function, such as acting as an oncogenic RBP sponge to inhibit cancer. In this study, we design a deep learning framework, CRPBsites, to predict the binding sites of RBPs on circRNAs. In this model, the sequences of variable-length binding sites are transformed into embedding vectors by word2vec model. Bidirectional LSTM is used to encode the embedding vectors of binding sites, and then they are fed into another LSTM decoder for decoding and classification tasks. To train and test the model, we construct four datasets that contain sequences of variable-length binding sites on circRNAs, and each set corresponds to an RBP, which is overexpressed in bladder cancer tissues. Experimental results on four datasets and comparison with other existing models show that CRPBsites has superior performance. Afterwards, we found that there were highly similar binding motifs in the four binding site datasets. Finally, we applied well-trained CRPBsites to identify the binding sites of IGF2BP1 on circCDYL, and the results proved the effectiveness of this method. In conclusion, CRPBsites is an effective prediction model for circRNA-RBP interaction site identification. We hope that CRPBsites can provide valuable guidance for experimental studies on the influence of circRNA on post-transcriptional regulation.

Description: Intratumoral heterogeneity is a well-documented feature of human cancers and is associated with outcome and treatment resistance. However, a heterogeneous tumor transcriptome contributes an unknown level of variability to analyses of differentially expressed genes (DEGs) that may contribute to phenotypes of interest, including treatment response. Although current clinical practice and the vast majority of research studies use a single sample from each patient, decreasing costs of sequencing technologies and computing power have made repeated-measures analyses increasingly economical. Repeatedly sampling the same tumor increases the statistical power of DEG analysis, which is indispensable toward downstream analysis and also increases one’s understanding of within-tumor variance, which may affect conclusions. Here, we compared five different methods for analyzing gene expression profiles derived from repeated sampling of human prostate tumors in two separate cohorts of patients. We also benchmarked the sensitivity of generalized linear models to linear mixed models for identifying DEGs contributing to relevant prostate cancer pathways based on a ground-truth model.

Description: Efforts to elucidate protein–DNA interactions at the molecular level rely in part on accurate predictions of DNA-binding residues in protein sequences. While there are over a dozen computational predictors of the DNA-binding residues, they are DNA-type agnostic and significantly cross-predict residues that interact with other ligands as DNA binding. We leverage a custom-designed machine learning architecture to introduce DNAgenie, first-of-its-kind predictor of residues that interact with A-DNA, B-DNA and single-stranded DNA. DNAgenie uses a comprehensive physiochemical profile extracted from an input protein sequence and implements a two-step refinement process to provide accurate predictions and to minimize the cross-predictions. Comparative tests on an independent test dataset demonstrate that DNAgenie outperforms the current methods that we adapt to predict residue-level interactions with the three DNA types. Further analysis finds that the use of the second (refinement) step leads to a substantial reduction in the cross predictions. Empirical tests show that DNAgenie’s outputs that are converted to coarse-grained protein-level predictions compare favorably against recent tools that predict which DNA-binding proteins interact with double-stranded versus single-stranded DNAs. Moreover, predictions from the sequences of the whole human proteome reveal that the results produced by DNAgenie substantially overlap with the known DNA-binding proteins while also including promising leads for several hundred previously unknown putative DNA binders. These results suggest that DNAgenie is a valuable tool for the sequence-based characterization of protein functions. The DNAgenie’s webserver is available at http://biomine.cs.vcu.edu/servers/DNAgenie/.

Description: Accurate prediction of immunogenic peptide recognized by T cell receptor (TCR) can greatly benefit vaccine development and cancer immunotherapy. However, identifying immunogenic peptides accurately is still a huge challenge. Most of the antigen peptides predicted in silico fail to elicit immune responses in vivo without considering TCR as a key factor. This inevitably causes costly and time-consuming experimental validation test for predicted antigens. Therefore, it is necessary to develop novel computational methods for precisely and effectively predicting immunogenic peptide recognized by TCR. Here, we described DLpTCR, a multimodal ensemble deep learning framework for predicting the likelihood of interaction between single/paired chain(s) of TCR and peptide presented by major histocompatibility complex molecules. To investigate the generality and robustness of the proposed model, COVID-19 data and IEDB data were constructed for independent evaluation. The DLpTCR model exhibited high predictive power with area under the curve up to 0.91 on COVID-19 data while predicting the interaction between peptide and single TCR chain. Additionally, the DLpTCR model achieved the overall accuracy of 81.03% on IEDB data while predicting the interaction between peptide and paired TCR chains. The results demonstrate that DLpTCR has the ability to learn general interaction rules and generalize to antigen peptide recognition by TCR. A user-friendly webserver is available at http://jianglab.org.cn/DLpTCR/. Additionally, a stand-alone software package that can be downloaded from https://github.com/jiangBiolab/DLpTCR.

Description: Over the past decade, genome-wide assays for chromatin interactions in single cells have enabled the study of individual nuclei at unprecedented resolution and throughput. Current chromosome conformation capture techniques survey contacts for up to tens of thousands of individual cells, improving our understanding of genome function in 3D. However, these methods recover a small fraction of all contacts in single cells, requiring specialised processing of sparse interactome data. In this review, we highlight recent advances in methods for the interpretation of single-cell genomic contacts. After discussing the strengths and limitations of these methods, we outline frontiers for future development in this rapidly moving field.

Description: Good knowledge of a peptide’s tertiary structure is important for understanding its function and its interactions with its biological targets. APPTEST is a novel computational protocol that employs a neural network architecture and simulated annealing methods for the prediction of peptide tertiary structure from the primary sequence. APPTEST works for both linear and cyclic peptides of 5–40 natural amino acids. APPTEST is computationally efficient, returning predicted structures within a number of minutes. APPTEST performance was evaluated on a set of 356 test peptides; the best structure predicted for each peptide deviated by an average of 1.9Å from its experimentally determined backbone conformation, and a native or near-native structure was predicted for 97% of the target sequences. A comparison of APPTEST performance with PEP-FOLD, PEPstrMOD and PepLook across benchmark datasets of short, long and cyclic peptides shows that on average APPTEST produces structures more native than the existing methods in all three categories. This innovative, cutting-edge peptide structure prediction method is available as an online web server at https://research.timmons.eu/apptest, facilitating in silico study and design of peptides by the wider research community.

Description: Deep generative models have been an upsurge in the deep learning community since they were proposed. These models are designed for generating new synthetic data including images, videos and texts by fitting the data approximate distributions. In the last few years, deep generative models have shown superior performance in drug discovery especially de novo molecular design. In this study, deep generative models are reviewed to witness the recent advances of de novo molecular design for drug discovery. In addition, we divide those models into two categories based on molecular representations in silico. Then these two classical types of models are reported in detail and discussed about both pros and cons. We also indicate the current challenges in deep generative models for de novo molecular design. De novo molecular design automatically is promising but a long road to be explored.

Description: DNA methylation may be regulated by genetic variants within a genomic region, referred to as methylation quantitative trait loci (mQTLs). The changes of methylation levels can further lead to alterations of gene expression, and influence the risk of various complex human diseases. Detecting mQTLs may provide insights into the underlying mechanism of how genotypic variations may influence the disease risk. In this article, we propose a methylation random field (MRF) method to detect mQTLs by testing the association between the methylation level of a CpG site and a set of genetic variants within a genomic region. The proposed MRF has two major advantages over existing approaches. First, it uses a beta distribution to characterize the bimodal and interval properties of the methylation trait at a CpG site. Second, it considers multiple common and rare genetic variants within a genomic region to identify mQTLs. Through simulations, we demonstrated that the MRF had improved power over other existing methods in detecting rare variants of relatively large effect, especially when the sample size is small. We further applied our method to a study of congenital heart defects with 83 cardiac tissue samples and identified two mQTL regions, MRPS10 and PSORS1C1, which were colocalized with expression QTL in cardiac tissue. In conclusion, the proposed MRF is a useful tool to identify novel mQTLs, especially for studies with limited sample sizes.

Description: Antimicrobial resistance (AMR) poses a threat to global public health. To mitigate the impacts of AMR, it is important to identify the molecular mechanisms of AMR and thereby determine optimal therapy as early as possible. Conventional machine learning-based drug-resistance analyses assume genetic variations to be homogeneous, thus not distinguishing between coding and intergenic sequences. In this study, we represent genetic data from Mycobacterium tuberculosis as a graph, and then adopt a deep graph learning method—heterogeneous graph attention network (‘HGAT–AMR’)—to predict anti-tuberculosis (TB) drug resistance. The HGAT–AMR model is able to accommodate incomplete phenotypic profiles, as well as provide ‘attention scores’ of genes and single nucleotide polymorphisms (SNPs) both at a population level and for individual samples. These scores encode the inputs, which the model is ‘paying attention to’ in making its drug resistance predictions. The results show that the proposed model generated the best area under the receiver operating characteristic (AUROC) for isoniazid and rifampicin (98.53 and 99.10%), the best sensitivity for three first-line drugs (94.91% for isoniazid, 96.60% for ethambutol and 90.63% for pyrazinamide), and maintained performance when the data were associated with incomplete phenotypes (i.e. for those isolates for which phenotypic data for some drugs were missing). We also demonstrate that the model successfully identifies genes and SNPs associated with drug resistance, mitigating the impact of resistance profile while considering particular drug resistance, which is consistent with domain knowledge.

Description: Protein engineering and design principles employing the 20 standard amino acids have been extensively used to achieve stable protein scaffolds and deliver their specific activities. Although this confers some advantages, it often restricts the sequence, chemical space, and ultimately the functional diversity of proteins. Moreover, although site-specific incorporation of non-natural amino acids (nnAAs) has been proven to be a valuable strategy in protein engineering and therapeutics development, its utility in the affinity-maturation of nanobodies is not fully explored. Besides, current experimental methods do not routinely employ nnAAs due to their enormous library size and infinite combinations. To address this, we have developed an integrated computational pipeline employing structure-based protein design methodologies, molecular dynamics simulations and free energy calculations, for the binding affinity prediction of an nnAA-incorporated nanobody toward its target and selection of potent binders. We show that by incorporating halogenated tyrosines, the affinity of 9G8 nanobody can be improved toward epidermal growth factor receptor (EGFR), a crucial cancer target. Surface plasmon resonance (SPR) assays showed that the binding of several 3-chloro-l-tyrosine (3MY)-incorporated nanobodies were improved up to 6-fold into a picomolar range, and the computationally estimated binding affinities shared a Pearson’s r of 0.87 with SPR results. The improved affinity was found to be due to enhanced van der Waals interactions of key 3MY-proximate nanobody residues with EGFR, and an overall increase in the nanobody’s structural stability. In conclusion, we show that our method can facilitate screening large libraries and predict potent site-specific nnAA-incorporated nanobody binders against crucial disease-targets.

Description: Over the past few years, meta-analysis has become popular among biomedical researchers for detecting biomarkers across multiple cohort studies with increased predictive power. Combining datasets from different sources increases sample size, thus overcoming the issue related to limited sample size from each individual study and boosting the predictive power. This leads to an increased likelihood of more accurately predicting differentially expressed genes/proteins or significant biomarkers underlying the biological condition of interest. Currently, several meta-analysis methods and tools exist, each having its own strengths and limitations. In this paper, we survey existing meta-analysis methods, and assess the performance of different methods based on results from different datasets as well as assessment from prior knowledge of each method. This provides a reference summary of meta-analysis models and tools, which helps to guide end-users on the choice of appropriate models or tools for given types of datasets and enables developers to consider current advances when planning the development of new meta-analysis models and more practical integrative tools.

Description: Emerging research shows that circular RNA (circRNA) plays a crucial role in the diagnosis, occurrence and prognosis of complex human diseases. Compared with traditional biological experiments, the computational method of fusing multi-source biological data to identify the association between circRNA and disease can effectively reduce cost and save time. Considering the limitations of existing computational models, we propose a semi-supervised generative adversarial network (GAN) model SGANRDA for predicting circRNA–disease association. This model first fused the natural language features of the circRNA sequence and the features of disease semantics, circRNA and disease Gaussian interaction profile kernel, and then used all circRNA–disease pairs to pre-train the GAN network, and fine-tune the network parameters through labeled samples. Finally, the extreme learning machine classifier is employed to obtain the prediction result. Compared with the previous supervision model, SGANRDA innovatively introduced circRNA sequences and utilized all the information of circRNA–disease pairs during the pre-training process. This step can increase the information content of the feature to some extent and reduce the impact of too few known associations on the model performance. SGANRDA obtained AUC scores of 0.9411 and 0.9223 in leave-one-out cross-validation and 5-fold cross-validation, respectively. Prediction results on the benchmark dataset show that SGANRDA outperforms other existing models. In addition, 25 of the top 30 circRNA–disease pairs with the highest scores of SGANRDA in case studies were verified by recent literature. These experimental results demonstrate that SGANRDA is a useful model to predict the circRNA–disease association and can provide reliable candidates for biological experiments.

Description: Metabolomics, the comprehensive study of the metabolome, and lipidomics—the large-scale study of pathways and networks of cellular lipids—are major driving forces in enabling personalized medicine. Complicated and error-prone data analysis still remains a bottleneck, however, especially for identifying novel metabolites. Comparing experimental mass spectra to curated databases containing reference spectra has been the gold standard for identification of compounds, but constructing such databases is a costly and time-demanding task. Many software applications try to circumvent this process by utilizing cutting-edge advances in computational methods—including quantum chemistry and machine learning—and simulate mass spectra by performing theoretical, so called in silico fragmentations of compounds. Other solutions concentrate directly on experimental spectra and try to identify structural properties by investigating reoccurring patterns and the relationships between them. The considerable progress made in the field allows recent approaches to provide valuable clues to expedite annotation of experimental mass spectra. This review sheds light on individual strengths and weaknesses of these tools, and attempts to evaluate them—especially in view of lipidomics, when considering complex mixtures found in biological samples as well as mass spectrometer inter-instrument variability.

Description: Proteins interact with each other to play critical roles in many biological processes in cells. Although promising, laboratory experiments usually suffer from the disadvantages of being time-consuming and labor-intensive. The results obtained are often not robust and considerably uncertain. Due recently to advances in high-throughput technologies, a large amount of proteomics data has been collected and this presents a significant opportunity and also a challenge to develop computational models to predict protein–protein interactions (PPIs) based on these data. In this paper, we present a comprehensive survey of the recent efforts that have been made towards the development of effective computational models for PPI prediction. The survey introduces the algorithms that can be used to learn computational models for predicting PPIs, and it classifies these models into different categories. To understand their relative merits, the paper discusses different validation schemes and metrics to evaluate the prediction performance. Biological databases that are commonly used in different experiments for performance comparison are also described and their use in a series of extensive experiments to compare different prediction models are discussed. Finally, we present some open issues in PPI prediction for future work. We explain how the performance of PPI prediction can be improved if these issues are effectively tackled.

Description: Coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a severe and rapidly evolving epidemic. Now, although a few drugs and vaccines have been proved for its treatment and prevention, little systematic comments are made to explain its susceptibility to humans. A few scattered studies used bioinformatics methods to explore the role of microRNA (miRNA) in COVID-19 infection. Combining these timely reports and previous studies about virus and miRNA, we comb through the available clues and seemingly make the perspective reasonable that the COVID-19 cleverly exploits the interplay between the small miRNA and other biomolecules to avoid being effectively recognized and attacked from host immune protection as well to deactivate functional genes that are crucial for immune system. In detail, SARS-CoV-2 can be regarded as a sponge to adsorb host immune-related miRNA, which forces host fall into dysfunction status of immune system. Besides, SARS-CoV-2 encodes its own miRNAs, which can enter host cell and are not perceived by the host’s immune system, subsequently targeting host function genes to cause illnesses. Therefore, this article presents a reasonable viewpoint that the miRNA-based interplays between the host and SARS-CoV-2 may be the primary cause that SARS-CoV-2 accesses and attacks the host cells.

Description:   Essential genes are critical for the growth and survival of any organism. The machine learning approach complements the experimental methods to minimize the resources required for essentiality assays. Previous studies revealed the need to discover relevant features that significantly classify essential genes, improve on the generalizability of prediction models across organisms, and construct a robust gold standard as the class label for the train data to enhance prediction. Findings also show that a significant limitation of the machine learning approach is predicting conditionally essential genes. The essentiality status of a gene can change due to a specific condition of the organism. This review examines various methods applied to essential gene prediction task, their strengths, limitations and the factors responsible for effective computational prediction of essential genes. We discussed categories of features and how they contribute to the classification performance of essentiality prediction models. Five categories of features, namely, gene sequence, protein sequence, network topology, homology and gene ontology-based features, were generated for Caenorhabditis elegans to perform a comparative analysis of their essentiality prediction capacity. Gene ontology-based feature category outperformed other categories of features majorly due to its high correlation with the genes’ biological functions. However, the topology feature category provided the highest discriminatory power making it more suitable for essentiality prediction. The major limiting factor of machine learning to predict essential genes conditionality is the unavailability of labeled data for interest conditions that can train a classifier. Therefore, cooperative machine learning could further exploit models that can perform well in conditional essentiality predictions. Short abstract Identification of essential genes is imperative because it provides an understanding of the core structure and function, accelerating drug targets’ discovery, among other functions. Recent studies have applied machine learning to complement the experimental identification of essential genes. However, several factors are limiting the performance of machine learning approaches. This review aims to present the standard procedure and resources available for predicting essential genes in organisms, and also highlight the factors responsible for the current limitation in using machine learning for conditional gene essentiality prediction. The choice of features and ML technique was identified as an important factor to predict essential genes effectively.

Description: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), causative agent of the coronavirus disease 2019 (COVID-19) pandemic, is thought to release its RNA genome at either the cell surface or within endosomes, the balance being dependent on spike protein stability, and the complement of receptors, co-receptors and proteases. To investigate possible mediators of pH-dependence, pKa calculations have been made on a set of structures for spike protein ectodomain and fragments from SARS-CoV-2 and other coronaviruses. Dominating a heat map of the aggregated predictions, three histidine residues in S2 are consistently predicted as destabilizing in pre-fusion (all three) and post-fusion (two of the three) structures. Other predicted features include the more moderate energetics of surface salt–bridge interactions and sidechain–mainchain interactions. Two aspartic acid residues in partially buried salt-bridges (D290–R273 and R355–D398) have pKas that are calculated to be elevated and destabilizing in more open forms of the spike trimer. These aspartic acids are most stabilized in a tightly closed conformation that has been observed when linoleic acid is bound, and which also affects the interactions of D614. The D614G mutation is known to modulate the balance of closed to open trimer. It is suggested that D398 in particular contributes to a pH-dependence of the open/closed equilibrium, potentially coupled to the effects of linoleic acid binding and D614G mutation, and possibly also A570D mutation. These observations are discussed in the context of SARS-CoV-2 infection, mutagenesis studies, and other human coronaviruses.

Description: In genome-wide mixed model association analysis, we stratified the genomic mixed model into two hierarchies to estimate genomic breeding values (GBVs) using the genomic best linear unbiased prediction and statistically infer the association of GBVs with each SNP using the generalized least square. The hierarchical mixed model (Hi-LMM) can correct confounders effectively with polygenic effects as residuals for association tests, preventing potential false-negative errors produced with genome-wide rapid association using mixed model and regression or an efficient mixed-model association expedited (EMMAX). Meanwhile, the Hi-LMM performs the same statistical power as the exact mixed model association and the same computing efficiency as EMMAX. When the GBVs have been estimated precisely, the Hi-LMM can detect more quantitative trait nucleotides (QTNs) than existing methods. Especially under the Hi-LMM framework, joint association analysis can be made straightforward to improve the statistical power of detecting QTNs.

Description: Typical clustering analysis for large-scale genomics data combines two unsupervised learning techniques: dimensionality reduction and clustering (DR-CL) methods. It has been demonstrated that transforming gene expression to pathway-level information can improve the robustness and interpretability of disease grouping results. This approach, referred to as biological knowledge-driven clustering (BK-CL) approach, is often neglected, due to a lack of tools enabling systematic comparisons with more established DR-based methods. Moreover, classic clustering metrics based on group separability tend to favor the DR-CL paradigm, which may increase the risk of identifying less actionable disease subtypes that have ambiguous biological and clinical explanations. Hence, there is a need for developing metrics that assess biological and clinical relevance. To facilitate the systematic analysis of BK-CL methods, we propose a computational protocol for quantitative analysis of clustering results derived from both DR-CL and BK-CL methods. Moreover, we propose a new BK-CL method that combines prior knowledge of disease relevant genes, network diffusion algorithms and gene set enrichment analysis to generate robust pathway-level information. Benchmarking studies were conducted to compare the grouping results from different DR-CL and BK-CL approaches with respect to standard clustering evaluation metrics, concordance with known subtypes, association with clinical outcomes and disease modules in co-expression networks of genes. No single approach dominated every metric, showing the importance multi-objective evaluation in clustering analysis. However, we demonstrated that, on gene expression data sets derived from TCGA samples, the BK-CL approach can find groupings that provide significant prognostic value in both breast and prostate cancers.

Description: A typical single-cell RNA sequencing (scRNA-seq) experiment will measure on the order of 20 000 transcripts and thousands, if not millions, of cells. The high dimensionality of such data presents serious complications for traditional data analysis methods and, as such, methods to reduce dimensionality play an integral role in many analysis pipelines. However, few studies have benchmarked the performance of these methods on scRNA-seq data, with existing comparisons assessing performance via downstream analysis accuracy measures, which may confound the interpretation of their results. Here, we present the most comprehensive benchmark of dimensionality reduction methods in scRNA-seq data to date, utilizing over 300 000 compute hours to assess the performance of over 25 000 low-dimension embeddings across 33 dimensionality reduction methods and 55 scRNA-seq datasets. We employ a simple, yet novel, approach, which does not rely on the results of downstream analyses. Internal validation measures (IVMs), traditionally used as an unsupervised method to assess clustering performance, are repurposed to measure how well-formed biological clusters are after dimensionality reduction. Performance was further evaluated over nearly 200 000 000 iterations of DBSCAN, a density-based clustering algorithm, showing that hyperparameter optimization using IVMs as the objective function leads to near-optimal clustering. Methods were also assessed on the extent to which they preserve the global structure of the data, and on their computational memory and time requirements across a large range of sample sizes. Our comprehensive benchmarking analysis provides a valuable resource for researchers and aims to guide best practice for dimensionality reduction in scRNA-seq analyses, and we highlight Latent Dirichlet Allocation and Potential of Heat-diffusion for Affinity-based Transition Embedding as high-performing algorithms.

Description: Genes do not function independently; rather, they interact with each other to fulfill their joint tasks. Identification of gene–gene interactions has been critically important in elucidating the molecular mechanisms responsible for the variation of a phenotype. Regression models are commonly used to model the interaction between two genes with a linear product term. The interaction effect of two genes can be linear or nonlinear, depending on the true nature of the data. When nonlinear interactions exist, the linear interaction model may not be able to detect such interactions; hence, it suffers from substantial power loss. While the true interaction mechanism (linear or nonlinear) is generally unknown in practice, it is critical to develop statistical methods that can be flexible to capture the underlying interaction mechanism without assuming a specific model assumption. In this study, we develop a mixed kernel function which combines both linear and Gaussian kernels with different weights to capture the linear or nonlinear interaction of two genes. Instead of optimizing the weight function, we propose a grid search strategy and use a Cauchy transformation of the P-values obtained under different weights to aggregate the P-values. We further extend the two-gene interaction model to a high-dimensional setup using a de-biased LASSO algorithm. Extensive simulation studies are conducted to verify the performance of the proposed method. Application to two case studies further demonstrates the utility of the model. Our method provides a flexible and computationally efficient tool for disentangling complex gene–gene interactions associated with complex traits.

Description: Motivation Empowered by advanced genomics discovery tools, recent biomedical research has produced a massive amount of genomic data on (post-)transcriptional regulations related to transcription factors, microRNAs, long non-coding RNAs, epigenetic modifications and genetic variations. Computational modeling, as an essential research method, has generated promising testable quantitative models that represent complex interplay among different gene regulatory mechanisms based on these data in many biological systems. However, given the dynamic changes of interactome in chaotic systems such as cancers, and the dramatic growth of heterogeneous data on this topic, such promise has encountered unprecedented challenges in terms of model complexity and scalability. In this study, we introduce a new integrative machine learning approach that can infer multifaceted gene regulations in cancers with a particular focus on microRNA regulation. In addition to new strategies for data integration and graphical model fusion, a supervised deep learning model was integrated to identify conditional microRNA-mRNA interactions across different cancer stages. Results In a case study of human breast cancer, we have identified distinct gene regulatory networks associated with four progressive stages. The subsequent functional analysis focusing on microRNA-mediated dysregulation across stages has revealed significant changes in major cancer hallmarks, as well as novel pathological signaling and metabolic processes, which shed light on microRNAs’ regulatory roles in breast cancer progression. We believe this integrative model can be a robust and effective discovery tool to understand key regulatory characteristics in complex biological systems. Availability http://sbbi-panda.unl.edu/pin/

Description: Drug combination therapy is a promising strategy to treat complex diseases such as cancer and infectious diseases. However, current knowledge of drug combination therapies, especially in cancer patients, is limited because of adverse drug effects, toxicity and cell line heterogeneity. Screening new drug combinations requires substantial efforts since considering all possible combinations between drugs is infeasible and expensive. Therefore, building computational approaches, particularly machine learning methods, could provide an effective strategy to overcome drug resistance and improve therapeutic efficacy. In this review, we group the state-of-the-art machine learning approaches to analyze personalized drug combination therapies into three categories and discuss each method in each category. We also present a short description of relevant databases used as a benchmark in drug combination therapies and provide a list of well-known, publicly available interactive data analysis portals. We highlight the importance of data integration on the identification of drug combinations. Finally, we address the advantages of combining multiple data sources on drug combination analysis by showing an experimental comparison.

Description: Accurate prediction of drug-target interactions (DTIs) through biological data can reduce the time and economic cost of drug development. The prediction method of DTIs based on a similarity network is attracting increasing attention. Currently, many studies have focused on predicting DTIs. However, such approaches do not consider the features of drugs and targets in multiple networks or how to extract and merge them. In this study, we proposed a Network EmbeDding framework in mulTiPlex networks (NEDTP) to predict DTIs. NEDTP builds a similarity network of nodes based on 15 heterogeneous information networks. Next, we applied a random walk to extract the topology information of each node in the network and learn it as a low-dimensional vector. Finally, the Gradient Boosting Decision Tree model was constructed to complete the classification task. NEDTP achieved accurate results in DTI prediction, showing clear advantages over several state-of-the-art algorithms. The prediction of new DTIs was also verified from multiple perspectives. In addition, this study also proposes a reasonable model for the widespread negative sampling problem of DTI prediction, contributing new ideas to future research. Code and data are available at https://github.com/LiangYu-Xidian/NEDTP.

Description: For epidemic prevention and control, the identification of SARS-CoV-2 subpopulations sharing similar micro-epidemiological patterns and evolutionary histories is necessary for a more targeted investigation into the links among COVID-19 outbreaks caused by SARS-CoV-2 with similar genetic backgrounds. Genomic sequencing analysis has demonstrated the ability to uncover viral genetic diversity. However, an objective analysis is necessary for the identification of SARS-CoV-2 subpopulations. Herein, we detected all the mutations in 186 682 SARS-CoV-2 isolates. We found that the GC content of the SARS-CoV-2 genome had evolved to be lower, which may be conducive to viral spread, and the frameshift mutation was rare in the global population. Next, we encoded the genomic mutations in binary form and used an unsupervised learning classifier, namely PhenoGraph, to classify this information. Consequently, PhenoGraph successfully identified 303 SARS-CoV-2 subpopulations, and we found that the PhenoGraph classification was consistent with, but more detailed and precise than the known GISAID clades (S, L, V, G, GH, GR, GV and O). By the change trend analysis, we found that the growth rate of SARS-CoV-2 diversity has slowed down significantly. We also analyzed the temporal, spatial and phylogenetic relationships among the subpopulations and revealed the evolutionary trajectory of SARS-CoV-2 to a certain extent. Hence, our results provide a better understanding of the patterns and trends in the genomic evolution and epidemiology of SARS-CoV-2.

Description: Neuropeptides acting as signaling molecules in the nervous system of various animals play crucial roles in a wide range of physiological functions and hormone regulation behaviors. Neuropeptides offer many opportunities for the discovery of new drugs and targets for the treatment of neurological diseases. In recent years, there have been several data-driven computational predictors of various types of bioactive peptides, but the relevant work about neuropeptides is little at present. In this work, we developed an interpretable stacking model, named NeuroPpred-Fuse, for the prediction of neuropeptides through fusing a variety of sequence-derived features and feature selection methods. Specifically, we used six types of sequence-derived features to encode the peptide sequences and then combined them. In the first layer, we ensembled three base classifiers and four feature selection algorithms, which select non-redundant important features complementarily. In the second layer, the output of the first layer was merged and fed into logistic regression (LR) classifier to train the model. Moreover, we analyzed the selected features and explained the feasibility of the selected features. Experimental results show that our model achieved 90.6% accuracy and 95.8% AUC on the independent test set, outperforming the state-of-the-art models. In addition, we exhibited the distribution of selected features by these tree models and compared the results on the training set to that on the test set. These results fully showed that our model has a certain generalization ability. Therefore, we expect that our model would provide important advances in the discovery of neuropeptides as new drugs for the treatment of neurological diseases.

Description: The computational identification of long non-coding RNAs (lncRNAs) is important to study lncRNAs and their functions. Despite the existence of many computation tools for lncRNA identification, to our knowledge, there is no systematic evaluation of these tools on common datasets and no consensus regarding their performance and the importance of the features used. To fill this gap, in this study, we assessed the performance of 17 tools on several common datasets. We also investigated the importance of the features used by the tools. We found that the deep learning-based tools have the best performance in terms of identifying lncRNAs, and the peptide features do not contribute much to the tool accuracy. Moreover, when the transcripts in a cell type were considered, the performance of all tools significantly dropped, and the deep learning-based tools were no longer as good as other tools. Our study will serve as an excellent starting point for selecting tools and features for lncRNA identification.

Description: Circular RNAs (circRNAs) are a class of single-stranded, covalently closed RNA molecules with a variety of biological functions. Studies have shown that circRNAs are involved in a variety of biological processes and play an important role in the development of various complex diseases, so the identification of circRNA-disease associations would contribute to the diagnosis and treatment of diseases. In this review, we summarize the discovery, classifications and functions of circRNAs and introduce four important diseases associated with circRNAs. Then, we list some significant and publicly accessible databases containing comprehensive annotation resources of circRNAs and experimentally validated circRNA-disease associations. Next, we introduce some state-of-the-art computational models for predicting novel circRNA-disease associations and divide them into two categories, namely network algorithm-based and machine learning-based models. Subsequently, several evaluation methods of prediction performance of these computational models are summarized. Finally, we analyze the advantages and disadvantages of different types of computational models and provide some suggestions to promote the development of circRNA-disease association identification from the perspective of the construction of new computational models and the accumulation of circRNA-related data.

Description: Artificial intelligence methods offer exciting new capabilities for the discovery of biological mechanisms from raw data because they are able to detect vastly more complex patterns of association that cannot be captured by classical statistical tests. Among these methods, deep neural networks are currently among the most advanced approaches and, in particular, convolutional neural networks (CNNs) have been shown to perform excellently for a variety of difficult tasks. Despite that applications of this type of networks to high-dimensional omics data and, most importantly, meaningful interpretation of the results returned from such models in a biomedical context remains an open problem. Here we present, an approach applying a CNN to nonimage data for feature selection. Our pipeline, DeepFeature, can both successfully transform omics data into a form that is optimal for fitting a CNN model and can also return sets of the most important genes used internally for computing predictions. Within the framework, the Snowfall compression algorithm is introduced to enable more elements in the fixed pixel framework, and region accumulation and element decoder is developed to find elements or genes from the class activation maps. In comparative tests for cancer type prediction task, DeepFeature simultaneously achieved superior predictive performance and better ability to discover key pathways and biological processes meaningful for this context. Capabilities offered by the proposed framework can enable the effective use of powerful deep learning methods to facilitate the discovery of causal mechanisms in high-dimensional biomedical data.

Description: The identification of protein–ligand interaction plays a key role in biochemical research and drug discovery. Although deep learning has recently shown great promise in discovering new drugs, there remains a gap between deep learning-based and experimental approaches. Here, we propose a novel framework, named AIMEE, integrating AI model and enzymological experiments, to identify inhibitors against 3CL protease of SARS-CoV-2 (Severe acute respiratory syndrome coronavirus 2), which has taken a significant toll on people across the globe. From a bioactive chemical library, we have conducted two rounds of experiments and identified six novel inhibitors with a hit rate of 29.41%, and four of them showed an IC50 value

Description: A central goal of precision oncology is to administer an optimal drug treatment to each cancer patient. A common preclinical approach to tackle this problem has been to characterize the tumors of patients at the molecular and drug response levels, and employ the resulting datasets for predictive in silico modeling (mostly using machine learning). Understanding how and why the different variants of these datasets are generated is an important component of this process. This review focuses on providing such introduction aimed at scientists with little previous exposure to this research area.

Description: We aimed to investigate the genetic mechanisms associated with coronavirus disease of 2019 (COVID-19) outcomes in the host and to evaluate the possible associations between smoking and drinking behavior and three COVID-19 outcomes: severe COVID-19, hospitalized COVID-19 and COVID-19 infection. We described the genomic loci and risk genes associated with the COVID-19 outcomes, followed by functional analyses of the risk genes. Then, a summary data-based Mendelian randomization (SMR) analysis, and a transcriptome-wide association study (TWAS) were performed for the severe COVID-19 dataset. A two-sample Mendelian randomization (MR) analysis was used to evaluate the causal associations between various measures of smoking and alcohol consumption and the COVID-19 outcomes. A total of 26 protein-coding genes, enriched in chemokine binding, cytokine binding and senescence-related functions, were associated with either severe COVID-19 or hospitalized COVID-19. The SMR and the TWAS analyses highlighted functional implications of some GWAS hits and identified seven novel genes for severe COVID-19, including CCR5, CCR5AS, IL10RB, TAC4, RMI1 and TNFSF15, some of which are targets of approved or experimental drugs. According to our studies, increasing consumption of cigarettes per day by 1 standard deviation is related to a 2.3-fold increase in susceptibility to severe COVID-19 and a 1.6-fold increase in COVID-19-induced hospitalization. Contrarily, no significant links were found between alcohol consumption or binary smoking status and COVID-19 outcomes. Our study revealed some novel COVID-19 related genes and suggested that genetic liability to smoking may quantitatively contribute to an increased risk for a severe course of COVID-19.

Description: The identification of structural variations (SVs) and viral integrations in circulating tumor DNA (ctDNA) is a key step in precision oncology that may assist clinicians in treatment selection and monitoring. However, due to the short fragment size of ctDNA, it is challenging to accurately detect low-frequency SVs or SVs involving complex junctions in ctDNA sequencing data. Here, we describe Aperture, a new fast SV caller that applies a unique strategy of $k$-mer-based searching, binary label–based breakpoint detection and candidate clustering to detect SVs and viral integrations with high sensitivity, especially when junctions span repetitive regions. Aperture also employs a barcode-based filter to ensure specificity. Compared with existing methods, Aperture exhibits superior sensitivity and specificity in simulated, reference and real data tests, especially at low dilutions. Additionally, Aperture is able to predict sites of viral integration and identify complex SVs involving novel insertions and repetitive sequences in real patient data. Aperture is freely available at https://github.com/liuhc8/Aperture.

Description: The goal of precision oncology is to tailor treatment for patients individually using the genomic profile of their tumors. Pharmacogenomics datasets such as cancer cell lines are among the most valuable resources for drug sensitivity prediction, a crucial task of precision oncology. Machine learning methods have been employed to predict drug sensitivity based on the multiple omics data available for large panels of cancer cell lines. However, there are no comprehensive guidelines on how to properly train and validate such machine learning models for drug sensitivity prediction. In this paper, we introduce a set of guidelines for different aspects of training gene expression-based predictors using cell line datasets. These guidelines provide extensive analysis of the generalization of drug sensitivity predictors and challenge many current practices in the community including the choice of training dataset and measure of drug sensitivity. The application of these guidelines in future studies will enable the development of more robust preclinical biomarkers.

Description: In this study, we proposed a deep learning (DL) model for classifying individuals from mixtures of DNA samples using 27 short tandem repeats and 94 single nucleotide polymorphisms obtained through massively parallel sequencing protocol. The model was trained/tested/validated with sequenced data from 6 individuals and then evaluated using mixtures from forensic DNA samples. The model successfully identified both the major and the minor contributors with 100% accuracy for 90 DNA mixtures, that were manually prepared by mixing sequence reads of 3 individuals at different ratios. Furthermore, the model identified 100% of the major contributors and 50–80% of the minor contributors in 20 two-sample external-mixed-samples at ratios of 1:39 and 1:9, respectively. To further demonstrate the versatility and applicability of the pipeline, we tested it on whole exome sequence data to classify subtypes of 20 breast cancer patients and achieved an area under curve of 0.85. Overall, we present, for the first time, a complete pipeline, including sequencing data processing steps and DL steps, that is applicable across different NGS platforms. We also introduced a sliding window approach, to overcome the sequence length variation problem of sequencing data, and demonstrate that it improves the model performance dramatically.

Description: The advent of large-scale biomedical data and computational algorithms provides new opportunities for drug repurposing and discovery. It is of great interest to find an appropriate data representation and modeling method to facilitate these studies. The anatomical therapeutic chemical (ATC) classification system, proposed by the World Health Organization (WHO), is an essential source of information for drug repurposing and discovery. Besides, computational methods are applied to predict drug ATC classification. We conducted a systematic review of ATC computational prediction studies and revealed the differences in data sets, data representation, algorithm approaches, and evaluation metrics. We then proposed a deep fusion learning (DFL) framework to optimize the ATC prediction model, namely DeepATC. The methods based on graph convolutional network, inferring biological network and multimodel attentive fusion network were applied in DeepATC to extract the molecular topological information and low-dimensional representation from the molecular graph and heterogeneous biological networks. The results indicated that DeepATC achieved superior model performance with area under the curve (AUC) value at 0.968. Furthermore, the DFL framework was performed for the transcriptome data–based ATC prediction, as well as another independent task that is significantly relevant to drug discovery, namely drug–target interaction. The DFL-based model achieved excellent performance in the above-extended validation task, suggesting that the idea of aggregating the heterogeneous biological network and node’s (molecule or protein) self-topological features will bring inspiration for broader drug repurposing and discovery research.

Description: Advances in the prediction of the inter-residue distance for a protein sequence have increased the accuracy to predict the correct folds of proteins with distance information. Here, we propose a distance-guided protein folding algorithm based on generalized descent direction, named GDDfold, which achieves effective structural perturbation and potential minimization in two stages. In the global stage, random-based direction is designed using evolutionary knowledge, which guides conformation population to cross potential barriers and explore conformational space rapidly in a large range. In the local stage, locally rugged potential landscape can be explored with the aid of conjugate-based direction integrated into a specific search strategy, which can improve the exploitation ability. GDDfold is tested on 347 proteins of a benchmark set, 24 template-free modeling (FM) approaches targets of CASP13 and 20 FM targets of CASP14. Results show that GDDfold correctly folds [template modeling (TM) score ≥ = 0.5] 316 out of 347 proteins, where 65 proteins have TM scores that are greater than 0.8, and significantly outperforms Rosetta-dist (distance-assisted fragment assembly method) and L-BFGSfold (distance geometry optimization method). On CASP FM targets, GDDfold is comparable with five state-of-the-art full-version methods, namely, Quark, RaptorX, Rosetta, MULTICOM and trRosetta in the CASP 13 and 14 server groups.

Description: Cell fate conversion by overexpressing defined factors is a powerful tool in regenerative medicine. However, identifying key factors for cell fate conversion requires laborious experimental efforts; thus, many of such conversions have not been achieved yet. Nevertheless, cell fate conversions found in many published studies were incomplete as the expression of important gene sets could not be manipulated thoroughly. Therefore, the identification of master transcription factors for complete and efficient conversion is crucial to render this technology more applicable clinically. In the past decade, systematic analyses on various single-cell and bulk OMICs data have uncovered numerous gene regulatory mechanisms, and made it possible to predict master gene regulators during cell fate conversion. By virtue of the sparse structure of master transcription factors and the group structure of their simultaneous regulatory effects on the cell fate conversion process, this study introduces a novel computational method predicting master transcription factors based on group sparse optimization technique integrating data from multi-OMICs levels, which can be applicable to both single-cell and bulk OMICs data with a high tolerance of data sparsity. When it is compared with current prediction methods by cross-referencing published and validated master transcription factors, it possesses superior performance. In short, this method facilitates fast identification of key regulators, give raise to the possibility of higher successful conversion rate and in the hope of reducing experimental cost.

Description: Recent advances in bioinformatics analyses have led to the development of novel tools enabling the capture and trajectory mapping of single-cell RNA sequencing (scRNAseq) data. However, there is a lack of methods to assess the contributions of biological pathways and transcription factors to an overall developmental trajectory mapped from scRNAseq data. In this manuscript, we present a simplified approach for trajectory inference of pathway significance (TIPS) that leverages existing knowledgebases of functional pathways and other gene lists to provide further mechanistic insights into a biological process. TIPS identifies key pathways which contribute to a process of interest, as well as the individual genes that best reflect these changes. TIPS also provides insight into the relative timing of pathway changes, as well as a suite of visualizations to enable simplified data interpretation of scRNAseq libraries generated using a wide range of techniques. The TIPS package can be run through either a web server or downloaded as a user-friendly GUI run in R, and may serve as a useful tool to help biologists perform deeper functional analyses and visualization of their single-cell data.

Description: Combination therapy has shown an obvious efficacy on complex diseases and can greatly reduce the development of drug resistance. However, even with high-throughput screens, experimental methods are insufficient to explore novel drug combinations. In order to reduce the search space of drug combinations, there is an urgent need to develop more efficient computational methods to predict novel drug combinations. In recent decades, more and more machine learning (ML) algorithms have been applied to improve the predictive performance. The object of this study is to introduce and discuss the recent applications of ML methods and the widely used databases in drug combination prediction. In this study, we first describe the concept and controversy of synergism between drug combinations. Then, we investigate various publicly available data resources and tools for prediction tasks. Next, ML methods including classic ML and deep learning methods applied in drug combination prediction are introduced. Finally, we summarize the challenges to ML methods in prediction tasks and provide a discussion on future work.

Description: In the past decade, convolutional neural networks (CNNs) have been used as powerful tools by scientists to solve visual data tasks. However, many efforts of convolutional neural networks in solving protein function prediction and extracting useful information from protein sequences have certain limitations. In this research, we propose a new method to improve the weaknesses of the previous method. mCNN-ETC is a deep learning model which can transform the protein evolutionary information into image-like data composed of 20 channels, which correspond to the 20 amino acids in the protein sequence. We constructed CNN layers with different scanning windows in parallel to enhance the useful pattern detection ability of the proposed model. Then we filtered specific patterns through the 1-max pooling layer before inputting them into the prediction layer. This research attempts to solve a basic problem in biology in terms of application: predicting electron transporters and classifying their corresponding complexes. The performance result reached an accuracy of 97.41%, which was nearly 6% higher than its predecessor. We have also published a web server on http://bio219.bioinfo.yzu.edu.tw, which can be used for research purposes free of charge.

Description: Clustering and cell type classification are a vital step of analyzing scRNA-seq data to reveal the complexity of the tissue (e.g. the number of cell types and the transcription characteristics of the respective cell type). Recently, deep learning-based single-cell clustering algorithms become popular since they integrate the dimensionality reduction with clustering. But these methods still have unstable clustering effects for the scRNA-seq datasets with high dropouts or noise. In this study, a novel single-cell RNA-seq deep embedding clustering via convolutional autoencoder embedding and soft K-means (scCAEs) is proposed by simultaneously learning the feature representation and clustering. It integrates the deep learning with convolutional autoencoder to characterize scRNA-seq data and proposes a regularized soft K-means algorithm to cluster cell populations in a learned latent space. Next, a novel constraint is introduced to the clustering objective function to iteratively optimize the clustering results, and more importantly, it is theoretically proved that this objective function optimization ensures the convergence. Moreover, it adds the reconstruction loss to the objective function combining the dimensionality reduction with clustering to find a more suitable embedding space for clustering. The proposed method is validated on a variety of datasets, in which the number of clusters in the mentioned datasets ranges from 4 to 46, and the number of cells ranges from 90 to 30 302. The experimental results show that scCAEs is superior to other state-of-the-art methods on the mentioned datasets, and it also keeps the satisfying compatibility and robustness. In addition, for single-cell datasets with the batch effects, scCAEs can ensure the cell separation while removing batch effects.

Description: It is becoming evident that holistic perspectives toward cancer are crucial in deciphering the overwhelming complexity of tumors. Single-layer analysis of genome-wide data has greatly contributed to our understanding of cellular systems and their perturbations. However, fundamental gaps in our knowledge persist and hamper the design of effective interventions. It is becoming more apparent than ever, that cancer should not only be viewed as a disease of the genome but as a disease of the cellular system. Integrative multilayer approaches are emerging as vigorous assets in our endeavors to achieve systemic views on cancer biology. Herein, we provide a comprehensive review of the approaches, methods and technologies that can serve to achieve systemic perspectives of cancer. We start with genome-wide single-layer approaches of omics analyses of cellular systems and move on to multilayer integrative approaches in which in-depth descriptions of proteogenomics and network-based data analysis are provided. Proteogenomics is a remarkable example of how the integration of multiple levels of information can reduce our blind spots and increase the accuracy and reliability of our interpretations and network-based data analysis is a major approach for data interpretation and a robust scaffold for data integration and modeling. Overall, this review aims to increase cross-field awareness of the approaches and challenges regarding the omics-based study of cancer and to facilitate the necessary shift toward holistic approaches.

Description: Graph is a natural data structure for describing complex systems, which contains a set of objects and relationships. Ubiquitous real-life biomedical problems can be modeled as graph analytics tasks. Machine learning, especially deep learning, succeeds in vast bioinformatics scenarios with data represented in Euclidean domain. However, rich relational information between biological elements is retained in the non-Euclidean biomedical graphs, which is not learning friendly to classic machine learning methods. Graph representation learning aims to embed graph into a low-dimensional space while preserving graph topology and node properties. It bridges biomedical graphs and modern machine learning methods and has recently raised widespread interest in both machine learning and bioinformatics communities. In this work, we summarize the advances of graph representation learning and its representative applications in bioinformatics. To provide a comprehensive and structured analysis and perspective, we first categorize and analyze both graph embedding methods (homogeneous graph embedding, heterogeneous graph embedding, attribute graph embedding) and graph neural networks. Furthermore, we summarize their representative applications from molecular level to genomics, pharmaceutical and healthcare systems level. Moreover, we provide open resource platforms and libraries for implementing these graph representation learning methods and discuss the challenges and opportunities of graph representation learning in bioinformatics. This work provides a comprehensive survey of emerging graph representation learning algorithms and their applications in bioinformatics. It is anticipated that it could bring valuable insights for researchers to contribute their knowledge to graph representation learning and future-oriented bioinformatics studies.

Description: Missing values are common in high-throughput mass spectrometry data. Two strategies are available to address missing values: (i) eliminate or impute the missing values and apply statistical methods that require complete data and (ii) use statistical methods that specifically account for missing values without imputation (imputation-free methods). This study reviews the effect of sample size and percentage of missing values on statistical inference for multiple methods under these two strategies. With increasing missingness, the ability of imputation and imputation-free methods to identify differentially and non-differentially regulated compounds in a two-group comparison study declined. Random forest and k-nearest neighbor imputation combined with a Wilcoxon test performed well in statistical testing for up to 50% missingness with little bias in estimating the effect size. Quantile regression imputation accompanied with a Wilcoxon test also had good statistical testing outcomes but substantially distorted the difference in means between groups. None of the imputation-free methods performed consistently better for statistical testing than imputation methods.

Description: Phage therapy has become one of the most promising alternatives to antibiotics in the treatment of bacterial diseases, and identifying phage-host interactions (PHIs) helps to understand the possible mechanism through which a phage infects bacteria to guide the development of phage therapy. Compared with wet experiments, computational methods of identifying PHIs can reduce costs and save time and are more effective and economic. In this paper, we propose a PHI prediction method with a generative adversarial network (GAN)-based data augmentation and sequence-based feature fusion (PHIAF). First, PHIAF applies a GAN-based data augmentation module, which generates pseudo PHIs to alleviate the data scarcity. Second, PHIAF fuses the features originated from DNA and protein sequences for better performance. Third, PHIAF utilizes an attention mechanism to consider different contributions of DNA/protein sequence-derived features, which also provides interpretability of the prediction model. In computational experiments, PHIAF outperforms other state-of-the-art PHI prediction methods when evaluated via 5-fold cross-validation (AUC and AUPR are 0.88 and 0.86, respectively). An ablation study shows that data augmentation, feature fusion and an attention mechanism are all beneficial to improve the prediction performance of PHIAF. Additionally, four new PHIs with the highest PHIAF score in the case study were verified by recent literature. In conclusion, PHIAF is a promising tool to accelerate the exploration of phage therapy.

Description: The emergence of single cell RNA sequencing has facilitated the studied of genomes, transcriptomes and proteomes. As available single-cell RNA-seq datasets are released continuously, one of the major challenges facing traditional RNA analysis tools is the high-dimensional, high-sparsity, high-noise and large-scale characteristics of single-cell RNA-seq data. Deep learning technologies match the characteristics of single-cell RNA-seq data perfectly and offer unprecedented promise. Here, we give a systematic review for most popular single-cell RNA-seq analysis methods and tools based on deep learning models, involving the procedures of data preprocessing (quality control, normalization, data correction, dimensionality reduction and data visualization) and clustering task for downstream analysis. We further evaluate the deep model-based analysis methods of data correction and clustering quantitatively on 11 gold standard datasets. Moreover, we discuss the data preferences of these methods and their limitations, and give some suggestions and guidance for users to select appropriate methods and tools.

Description: Chemosensitivity assays are commonly used for preclinical drug discovery and clinical trial optimization. However, data from independent assays are often discordant, largely attributed to uncharacterized variation in the experimental materials and protocols. We report here the launching of Minimal Information for Chemosensitivity Assays (MICHA), accessed via https://micha-protocol.org. Distinguished from existing efforts that are often lacking support from data integration tools, MICHA can automatically extract publicly available information to facilitate the assay annotation including: 1) compounds, 2) samples, 3) reagents and 4) data processing methods. For example, MICHA provides an integrative web server and database to obtain compound annotation including chemical structures, targets and disease indications. In addition, the annotation of cell line samples, assay protocols and literature references can be greatly eased by retrieving manually curated catalogues. Once the annotation is complete, MICHA can export a report that conforms to the FAIR principle (Findable, Accessible, Interoperable and Reusable) of drug screening studies. To consolidate the utility of MICHA, we provide FAIRified protocols from five major cancer drug screening studies as well as six recently conducted COVID-19 studies. With the MICHA web server and database, we envisage a wider adoption of a community-driven effort to improve the open access of drug sensitivity assays.

Description: Real-world evaluations of metagenomic reconstructions are challenged by distinguishing reconstruction artifacts from genes and proteins present in situ. Here, we evaluate short-read-only, long-read-only and hybrid assembly approaches on four different metagenomic samples of varying complexity. We demonstrate how different assembly approaches affect gene and protein inference, which is particularly relevant for downstream functional analyses. For a human gut microbiome sample, we use complementary metatranscriptomic and metaproteomic data to assess the metagenomic data-based protein predictions. Our findings pave the way for critical assessments of metagenomic reconstructions. We propose a reference-independent solution, which exploits the synergistic effects of multi-omic data integration for the in situ study of microbiomes using long-read sequencing data.

Description: Continuous evaluation of drug safety is needed following approval to determine adverse events (AEs) in patient populations with diverse backgrounds. Spontaneous reporting systems are an important source of information for the detection of AEs not identified in clinical trials and for safety assessments that reflect the real-world use of drugs in specific populations and clinical settings. The use of spontaneous reporting systems is expected to detect drug-related AEs early after the launch of a new drug. Spontaneous reporting systems do not contain data on the total number of patients that use a drug; therefore, signal detection by disproportionality analysis, focusing on differences in the ratio of AE reports, is frequently used. In recent years, new analyses have been devised, including signal detection methods focused on the difference in the time to onset of an AE, methods that consider the patient background and those that identify drug–drug interactions. However, unlike commonly used statistics, the results of these analyses are open to misinterpretation if the method and the characteristics of the spontaneous reporting system cannot be evaluated properly. Therefore, this review describes signal detection using data mining, considering traditional methods and the latest knowledge, and their limitations.

Description: DNA N6-methyladenine is an important type of DNA modification that plays important roles in multiple biological processes. Despite the recent progress in developing DNA 6mA site prediction methods, several challenges remain to be addressed. For example, although the hand-crafted features are interpretable, they contain redundant information that may bias the model training and have a negative impact on the trained model. Furthermore, although deep learning (DL)-based models can perform feature extraction and classification automatically, they lack the interpretability of the crucial features learned by those models. As such, considerable research efforts have been focused on achieving the trade-off between the interpretability and straightforwardness of DL neural networks. In this study, we develop two new DL-based models for improving the prediction of N6-methyladenine sites, termed LA6mA and AL6mA, which use bidirectional long short-term memory to respectively capture the long-range information and self-attention mechanism to extract the key position information from DNA sequences. The performance of the two proposed methods is benchmarked and evaluated on the two model organisms Arabidopsis thaliana and Drosophila melanogaster. On the two benchmark datasets, LA6mA achieves an area under the receiver operating characteristic curve (AUROC) value of 0.962 and 0.966, whereas AL6mA achieves an AUROC value of 0.945 and 0.941, respectively. Moreover, an in-depth analysis of the attention matrix is conducted to interpret the important information, which is hidden in the sequence and relevant for 6mA site prediction. The two novel pipelines developed for DNA 6mA site prediction in this work will facilitate a better understanding of the underlying principle of DL-based DNA methylation site prediction and its future applications.

Description: In genome-wide association studies (GWAS), it has become commonplace to test millions of single-nucleotide polymorphisms (SNPs) for phenotypic association. Gene-based testing can improve power to detect weak signal by reducing multiple testing and pooling signal strength. While such tests account for linkage disequilibrium (LD) structure of SNP alleles within each gene, current approaches do not capture LD of SNPs falling in different nearby genes, which can induce correlation of gene-based test statistics. We introduce an algorithm to account for this correlation. When a gene’s test statistic is independent of others, it is assessed separately; when test statistics for nearby genes are strongly correlated, their SNPs are agglomerated and tested as a locus. To provide insight into SNPs and genes driving association within loci, we develop an interactive visualization tool to explore localized signal. We demonstrate our approach in the context of weakly powered GWAS for autism spectrum disorder, which is contrasted to more highly powered GWAS for schizophrenia and educational attainment. To increase power for these analyses, especially those for autism, we use adaptive $P$-value thresholding, guided by high-dimensional metadata modeled with gradient boosted trees, highlighting when and how it can be most useful. Notably our workflow is based on summary statistics.

Description: Therapeutic peptides are important for understanding the correlation between peptides and their therapeutic diagnostic potential. The therapeutic peptides can be further divided into different types based on therapeutic function sharing different characteristics. Although some computational approaches have been proposed to predict different types of therapeutic peptides, they failed to accurately predict all types of therapeutic peptides. In this study, a predictor called PreTP-EL has been proposed via employing the ensemble learning approach to fuse the different features and machine learning techniques in order to capture the different characteristics of various therapeutic peptides. Experimental results showed that PreTP-EL outperformed other competing methods. Availability and implementation: A user-friendly web-server of PreTP-EL predictor is available at http://bliulab.net/PreTP-EL.

Description: Deep residual learning has shown great success in protein contact prediction. In this study, a new deep residual learning-based protein contact prediction model was developed. Comparing with previous models, a new type of residual block hybridizing 1D and 2D convolutions was designed to increase the effective receptive field of the residual network, and a new loss function emphasizing the easily misclassified residue pairs was proposed to enhance the model training. The developed protein contact prediction model referred to as DRN-1D2D was first evaluated on 105 CASP11 targets, 76 CAMEO hard targets and 398 membrane proteins together with two in house-developed reference models based on either the standard 2D residual block or the traditional BCE loss function, from which we confirmed that both the dimensional hybrid residual block and the singularity enhanced loss function can be employed to improve the model performance for protein contact prediction. DRN-1D2D was further evaluated on 39 CASP13 and CASP14 free modeling targets together with the two reference models and six state-of-the-art protein contact prediction models including DeepCov, DeepCon, DeepConPred2, SPOT-Contact, RaptorX-Contact and TripleRes. The result shows that DRN-1D2D consistently achieved the best performance among all these models.

Description: Several prediction algorithms and tools have been developed in the last two decades to predict protein and peptide aggregation. These in silico tools aid to predict the aggregation propensity and amyloidogenicity as well as the identification of aggregation-prone regions. Despite the immense interest in the field, it is of prime importance to systematically compare these algorithms for their performance. In this review, we have provided a rigorous performance analysis of nine prediction tools using a variety of assessments. The assessments were carried out on several non-redundant datasets ranging from hexapeptides to protein sequences as well as amyloidogenic antibody light chains to soluble protein sequences. Our analysis reveals the robustness of the current prediction tools and the scope for improvement in their predictive performances. Insights gained from this work provide critical guidance to the scientific community on advantages and limitations of different aggregation prediction methods and make informed decisions about their research needs.

Description: To enable personalized cancer treatment, machine learning models have been developed to predict drug response as a function of tumor and drug features. However, most algorithm development efforts have relied on cross-validation within a single study to assess model accuracy. While an essential first step, cross-validation within a biological data set typically provides an overly optimistic estimate of the prediction performance on independent test sets. To provide a more rigorous assessment of model generalizability between different studies, we use machine learning to analyze five publicly available cell line-based data sets: National Cancer Institute 60, ancer Therapeutics Response Portal (CTRP), Genomics of Drug Sensitivity in Cancer, Cancer Cell Line Encyclopedia and Genentech Cell Line Screening Initiative (gCSI). Based on observed experimental variability across studies, we explore estimates of prediction upper bounds. We report performance results of a variety of machine learning models, with a multitasking deep neural network achieving the best cross-study generalizability. By multiple measures, models trained on CTRP yield the most accurate predictions on the remaining testing data, and gCSI is the most predictable among the cell line data sets included in this study. With these experiments and further simulations on partial data, two lessons emerge: (1) differences in viability assays can limit model generalizability across studies and (2) drug diversity, more than tumor diversity, is crucial for raising model generalizability in preclinical screening.

Description: Since the outbreak of SARS-CoV-2, the etiologic agent of the COVID-19 pandemic, the viral genome has acquired numerous mutations with the potential to alter the viral infectivity and antigenicity. Part of mutations in SARS-CoV-2 spike protein has conferred virus the ability to spread more quickly and escape from the immune response caused by the monoclonal neutralizing antibody or vaccination. Herein, we summarize the spatiotemporal distribution of mutations in spike protein, and present recent efforts and progress in investigating the impacts of those mutations on viral infectivity and antigenicity. As mutations continue to emerge in SARS-CoV-2, we strive to provide systematic evaluation of mutations in spike protein, which is vitally important for the subsequent improvement of vaccine and therapeutic neutralizing antibody strategies.

Description: The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic has triggered an unprecedented international effort to sequence complete viral genomes. We leveraged this wealth of information to characterize the substitution spectrum of SARS-CoV-2 and to compare it with those of other human and animal coronaviruses. We show that, once nucleotide composition is taken into account, human and most animal coronaviruses display a mutation spectrum dominated by C to U and G to U substitutions, a feature that is not shared by other positive-sense RNA viruses. However, the proportions of C to U and G to U substitutions tend to decrease as divergence increases, suggesting that, whatever their origin, a proportion of these changes is subsequently eliminated by purifying selection. Analysis of the sequence context of C to U substitutions showed little evidence of apolipoprotein B mRNA editing catalytic polypeptide-like (APOBEC)-mediated editing and such contexts were similar for SARS-CoV-2 and Middle East respiratory syndrome coronavirus sampled from different hosts, despite different repertoires of APOBEC3 proteins in distinct species. Conversely, we found evidence that C to U and G to U changes affect CpG dinucleotides at a frequency higher than expected. Whereas this suggests ongoing selective reduction of CpGs, this effect alone cannot account for the substitution spectra. Finally, we show that, during the first months of SARS-CoV-2 pandemic spread, the frequency of both G to U and C to U substitutions increased. Our data suggest that the substitution spectrum of SARS-CoV-2 is determined by an interplay of factors, including intrinsic biases of the replication process, avoidance of CpG dinucleotides and other constraints exerted by the new host.

Description: Drug combination is a sensible strategy for disease treatment because it improves the treatment efficacy and reduces concomitant side effects. Due to the large number of possible combinations among candidate compounds, exhaustive screening is prohibitive. Currently, a large number of studies have focused on predicting potential drug combinations. However, these methods are not entirely satisfactory in terms of performance and scalability. In this paper, we proposed a Network Embedding frameWork in MultIplex Network (NEWMIN) to predict synthetic drug combinations. Based on a multiplex drug similarity network, we offered alternative methods to integrate useful information from different aspects and to decide the quantitative importance of each network. For drug combination prediction, we found seven novel drug combinations that have been validated by external sources among the top-ranked predictions of our model. To verify the feasibility of NEWMIN, we compared NEWMIN with other five methods, for which it showed better performance than other methods in terms of the area under the precision-recall curve and receiver operating characteristic curve.

Description: A comprehensive analysis of omics data can require vast computational resources and access to varied data sources that must be integrated into complex, multi-step analysis pipelines. Execution of many such analyses can be accelerated by applying the cloud computing paradigm, which provides scalable resources for storing data of different types and parallelizing data analysis computations. Moreover, these resources can be reused for different multi-omics analysis scenarios. Traditionally, developers are required to manage a cloud platform’s underlying infrastructure, configuration, maintenance and capacity planning. The serverless computing paradigm simplifies these operations by automatically allocating and maintaining both servers and virtual machines, as required for analysis tasks. This paradigm offers highly parallel execution and high scalability without manual management of the underlying infrastructure, freeing developers to focus on operational logic. This paper reviews serverless solutions in bioinformatics and evaluates their usage in omics data analysis and integration. We start by reviewing the application of the cloud computing model to a multi-omics data analysis and exposing some shortcomings of the early approaches. We then introduce the serverless computing paradigm and show its applicability for performing an integrative analysis of multiple omics data sources in the context of the COVID-19 pandemic.

Description: The cooperativity of transcription factors (TFs) is a widespread phenomenon in the gene regulation system. However, the interaction patterns between TF binding motifs remain elusive. The recent high-throughput assays, CAP-SELEX, have identified over 600 composite DNA sites (i.e. heterodimeric motifs) bound by cooperative TF pairs. However, there are over 25 000 inferentially effective heterodimeric TFs in the human cells. It is not practically feasible to validate all heterodimeric motifs due to cost and labor. We introduce DeepMotifSyn, a deep learning-based tool for synthesizing heterodimeric motifs from monomeric motif pairs. Specifically, DeepMotifSyn is composed of heterodimeric motif generator and evaluator. The generator is a U-Net-based neural network that can synthesize heterodimeric motifs from aligned motif pairs. The evaluator is a machine learning-based model that can score the generated heterodimeric motif candidates based on the motif sequence features. Systematic evaluations on CAP-SELEX data illustrate that DeepMotifSyn significantly outperforms the current state-of-the-art predictors. In addition, DeepMotifSyn can synthesize multiple heterodimeric motifs with different orientation and spacing settings. Such a feature can address the shortcomings of previous models. We believe DeepMotifSyn is a more practical and reliable model than current predictors on heterodimeric motif synthesis. Contact:  kc.w@cityu.edu.hk

Description: Whether risk genes of severe coronavirus disease 2019 (COVID-19) from genome-wide association study could play their regulatory roles by interacting with host genes that were interacted with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) proteins was worthy of exploration. In this study, we implemented a network-based approach by developing a user-friendly software Network Calculator (https://github.com/Haoxiang-Qi/Network-Calculator.git). By using Network Calculator, we identified a network composed of 13 risk genes and 28 SARS-CoV-2 interacted host genes that had the highest network proximity with each other, with a hub gene HNRNPK identified. Among these genes, 14 of them were identified to be differentially expressed in RNA-seq data from severe COVID-19 cases. Besides, by expression enrichment analysis in single-cell RNA-seq data, compared with mild COVID-19, these genes were significantly enriched in macrophage, T cell and epithelial cell for severe COVID-19. Meanwhile, 74 pathways were significantly enriched. Our analysis provided insights for the underlying genetic etiology of severe COVID-19 from the perspective of network biology.

Description: Protein post-translational modification (PTM) is an important regulatory mechanism that plays a key role in both normal and disease states. Acetylation on lysine residues is one of the most potent PTMs owing to its critical role in cellular metabolism and regulatory processes. Identifying protein lysine acetylation (Kace) sites is a challenging task in bioinformatics. To date, several machine learning-based methods for the in silico identification of Kace sites have been developed. Of those, a few are prokaryotic species-specific. Despite their attractive advantages and performances, these methods have certain limitations. Therefore, this study proposes a novel predictor STALLION (STacking-based Predictor for ProkAryotic Lysine AcetyLatION), containing six prokaryotic species-specific models to identify Kace sites accurately. To extract crucial patterns around Kace sites, we employed 11 different encodings representing three different characteristics. Subsequently, a systematic and rigorous feature selection approach was employed to identify the optimal feature set independently for five tree-based ensemble algorithms and built their respective baseline model for each species. Finally, the predicted values from baseline models were utilized and trained with an appropriate classifier using the stacking strategy to develop STALLION. Comparative benchmarking experiments showed that STALLION significantly outperformed existing predictor on independent tests. To expedite direct accessibility to the STALLION models, a user-friendly online predictor was implemented, which is available at: http://thegleelab.org/STALLION.

Description: Motivation The Estimation of Model Accuracy problem is a cornerstone problem in the field of Bioinformatics. As of CASP14, there are 79 global QA methods, and a minority of 39 residue-level QA methods with very few of them working on protein complexes. Here, we introduce ZoomQA, a novel, single-model method for assessing the accuracy of a tertiary protein structure/complex prediction at residue level, which have many applications such as drug discovery. ZoomQA differs from others by considering the change in chemical and physical features of a fragment structure (a portion of a protein within a radius $r$ of the target amino acid) as the radius of contact increases. Fourteen physical and chemical properties of amino acids are used to build a comprehensive representation of every residue within a protein and grade their placement within the protein as a whole. Moreover, we have shown the potential of ZoomQA to identify problematic regions of the SARS-CoV-2 protein complex. Results We benchmark ZoomQA on CASP14, and it outperforms other state-of-the-art local QA methods and rivals state of the art QA methods in global prediction metrics. Our experiment shows the efficacy of these new features and shows that our method is able to match the performance of other state-of-the-art methods without the use of homology searching against databases or PSSM matrices. Availability http://zoomQA.renzhitech.com

Description: Ferroptosis is a form of regulated cell death initiated by oxidative perturbations that can be blocked by iron chelators and lipophilic antioxidants, and ferroptosis may be the silver bullet treatment for multiple cancers, including immunotherapy- and chemotherapy-insensitive cancers such as pancreatic ductal adenocarcinoma (PDAC). Numerous studies have noted that long non-coding RNAs (lncRNAs) regulate the biological behaviour of cancer cells by binding to DNA, RNA and protein. However, few studies have reported the role of lncRNAs in ferroptosis processes and the function of ferroptosis-associated lncRNAs. The primary objective of the present study was to identify ferroptosis-related lncRNAs using bioinformatic approaches combined with experimental validation. The second objective was to construct a prognostic model to predict the overall survival of patients with PDAC. The present study identified ferroptosis-related lncRNAs using a bioinformatic approach and validated them in an independent pancreatic cancer cohort from Fudan University Shanghai Cancer Center. The lncRNA SLCO4A1-AS1 was identified as a novel molecule mediating ferroptosis resistance in vitro. A novel algorithm was used to construct a ‘0 or 1’ matrix-based prognosis model, which showed promising diagnostic accuracy for potential clinical translation (area under the curve = 0.89 for the 2-year survival rate). Notably, molecular subtypes classified by the risk scores of the model did not belong to any previously reported subtypes of PDAC. The immune microenvironment, metabolic activities, mutation landscape and ferroptosis sensitivity were significantly distinct between patients with different risk scores. Sensitivity (IC50) to 30 common anticancer drugs was analysed between patients with different risks, and imatinib and axitinib were found to be potential drugs for the treatment of patients with lower risk scores. Overall, we developed an accurate prognostic model based on the expression patterns of ferroptosis lncRNAs, which may contribute greatly to the evaluation of patient prognosis, molecular characteristics and treatment modalities and could be further translated into clinical applications.

Description: The development of single-cell ribonucleic acid (RNA) sequencing (scRNA-seq) technology has led to great opportunities for the identification of heterogeneous cell types in complex tissues. Clustering algorithms are of great importance to effectively identify different cell types. In addition, the definition of the distance between each two cells is a critical step for most clustering algorithms. In this study, we found that different distance measures have considerably different effects on clustering algorithms. Moreover, there is no specific distance measure that is applicable to all datasets. In this study, we introduce a new single-cell clustering method called SD-h, which generates an applicable distance measure for different kinds of datasets by optimally synthesizing commonly used distance measures. Then, hierarchical clustering is performed based on the new distance measure for more accurate cell-type clustering. SD-h was tested on nine frequently used scRNA-seq datasets and it showed great superiority over almost all the compared leading single-cell clustering algorithms.

Description: Although the current coronavirus disease 2019 (COVID-19) vaccines have been used worldwide to halt spread of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the emergence of new SARS-CoV-2 variants with E484K mutation shows significant resistance to the neutralization of vaccine sera. To better understand the resistant mechanism, we calculated the binding affinities of 26 antibodies to wild-type (WT) spike protein and to the protein harboring E484K mutation, respectively. The results showed that most antibodies (~85%) have weaker binding affinities to the E484K mutated spike protein than to the WT, indicating the high risk of immune evasion of the mutated virus from most of current antibodies. Binding free energy decomposition revealed that the residue E484 forms attraction with most antibodies, while the K484 has repulsion from most antibodies, which should be the main reason of the weaker binding affinities of E484K mutant to most antibodies. Impressively, a monoclonal antibody (mAb) combination was found to have much stronger binding affinity with E484K mutant than WT, which may work well against the mutated virus. Based on binding free energy decomposition, we predicted that the mutation of four more residues on receptor-binding domain (RBD) of spike protein, viz., F490, V483, G485 and S494, may have high risk of immune evasion, which we should pay close attention on during the development of new mAb therapeutics.

Description: Quantitative trait locus (QTL) analyses of multiomic molecular traits, such as gene transcription (eQTL), DNA methylation (mQTL) and histone modification (haQTL), have been widely used to infer the functional effects of genome variants. However, the QTL discovery is largely restricted by the limited study sample size, which demands higher threshold of minor allele frequency and then causes heavy missing molecular trait–variant associations. This happens prominently in single-cell level molecular QTL studies because of sample availability and cost. It is urgent to propose a method to solve this problem in order to enhance discoveries of current molecular QTL studies with small sample size. In this study, we presented an efficient computational framework called xQTLImp to impute missing molecular QTL associations. In the local-region imputation, xQTLImp uses multivariate Gaussian model to impute the missing associations by leveraging known association statistics of variants and the linkage disequilibrium (LD) around. In the genome-wide imputation, novel procedures are implemented to improve efficiency, including dynamically constructing a reused LD buffer, adopting multiple heuristic strategies and parallel computing. Experiments on various multiomic bulk and single-cell sequencing-based QTL datasets have demonstrated high imputation accuracy and novel QTL discovery ability of xQTLImp. Finally, a C++ software package is freely available at https://github.com/stormlovetao/QTLIMP.

Description: Single-cell RNA sequencing (scRNA-seq) technologies have been heavily developed to probe gene expression profiles at single-cell resolution. Deep imputation methods have been proposed to address the related computational challenges (e.g. the gene sparsity in single-cell data). In particular, the neural architectures of those deep imputation models have been proven to be critical for performance. However, deep imputation architectures are difficult to design and tune for those without rich knowledge of deep neural networks and scRNA-seq. Therefore, Surrogate-assisted Evolutionary Deep Imputation Model (SEDIM) is proposed to automatically design the architectures of deep neural networks for imputing gene expression levels in scRNA-seq data without any manual tuning. Moreover, the proposed SEDIM constructs an offline surrogate model, which can accelerate the computational efficiency of the architectural search. Comprehensive studies show that SEDIM significantly improves the imputation and clustering performance compared with other benchmark methods. In addition, we also extensively explore the performance of SEDIM in other contexts and platforms including mass cytometry and metabolic profiling in a comprehensive manner. Marker gene detection, gene ontology enrichment and pathological analysis are conducted to provide novel insights into cell-type identification and the underlying mechanisms. The source code is available at https://github.com/li-shaochuan/SEDIM.

Description: After experiencing the COVID-19 pandemic, it is widely acknowledged that a rapid drug repurposing method is highly needed. A series of useful drug repurposing tools have been developed based on data-driven modeling and network pharmacology. Based on the disease module, we identified several hub proteins that play important roles in the onset and development of the COVID-19, which are potential targets for repositioning approved drugs. Moreover, different network distance metrics were applied to quantify the relationship between drug targets and COVID-19 disease targets in the protein–protein-interaction (PPI) network and predict COVID-19 therapeutic effects of bioactive herbal ingredients and chemicals. Furthermore, the tentative mechanisms of candidates were illustrated through molecular docking and gene enrichment analysis. We obtained 15 chemical and 15 herbal ingredient candidates and found that different drugs may play different roles in the process of virus invasion and the onset and development of the COVID-19 disease. Given pandemic outbreaks, our method has an undeniable immense advantage in the feasibility analysis of drug repurposing or drug screening, especially in the analysis of herbal ingredients.

Description: Next-generation sequencing expands the known phage genomes rapidly. Unlike culture-based methods, the hosts of phages discovered from next-generation sequencing data remain uncharacterized. The high diversity of the phage genomes makes the host assignment task challenging. To solve the issue, we proposed a phage host prediction tool—DeepHost. To encode the phage genomes into matrices, we design a genome encoding method that applied various spaced $k$-mer pairs to tolerate sequence variations, including insertion, deletions, and mutations. DeepHost applies a convolutional neural network to predict host taxonomies. DeepHost achieves the prediction accuracy of 96.05% at the genus level (72 taxonomies) and 90.78% at the species level (118 taxonomies), which outperforms the existing phage host prediction tools by 10.16–30.48% and achieves comparable results to BLAST. For the genomes without hits in BLAST, DeepHost obtains the accuracy of 38.00% at the genus level and 26.47% at the species level, making it suitable for genomes of less homologous sequences with the existing datasets. DeepHost is alignment-free, and it is faster than BLAST, especially for large datasets. DeepHost is available at https://github.com/deepomicslab/DeepHost.

Description: Single-cell Hi-C data are a common data source for studying the differences in the three-dimensional structure of cell chromosomes. The development of single-cell Hi-C technology makes it possible to obtain batches of single-cell Hi-C data. How to quickly and effectively discriminate cell types has become one hot research field. However, the existing computational methods to predict cell types based on Hi-C data are found to be low in accuracy. Therefore, we propose a high accuracy cell classification algorithm, called scHiCStackL, based on single-cell Hi-C data. In our work, we first improve the existing data preprocessing method for single-cell Hi-C data, which allows the generated cell embedding better to represent cells. Then, we construct a two-layer stacking ensemble model for classifying cells. Experimental results show that the cell embedding generated by our data preprocessing method increases by 0.23, 1.22, 1.46 and 1.61$\%$ comparing with the cell embedding generated by the previously published method scHiCluster, in terms of the Acc, MCC, F1 and Precision confidence intervals, respectively, on the task of classifying human cells in the ML1 and ML3 datasets. When using the two-layer stacking ensemble framework with the cell embedding, scHiCStackL improves by 13.33, 19, 19.27 and 14.5 over the scHiCluster, in terms of the Acc, ARI, NMI and F1 confidence intervals, respectively. In summary, scHiCStackL achieves superior performance in predicting cell types using the single-cell Hi-C data. The webserver and source code of scHiCStackL are freely available at http://hww.sdu.edu.cn:8002/scHiCStackL/ and https://github.com/HaoWuLab-Bioinformatics/scHiCStackL, respectively.

Description: With the rapid development of single-cell sequencing techniques, several large-scale cell atlas projects have been launched across the world. However, it is still challenging to integrate single-cell RNA-seq (scRNA-seq) datasets with diverse tissue sources, developmental stages and/or few overlaps, due to the ambiguity in determining the batch information, which is particularly important for current batch-effect correction methods. Here, we present SCORE, a simple network-based integration methodology, which incorporates curated molecular network features to infer cellular states and generate a unified workflow for integrating scRNA-seq datasets. Validating on real single-cell datasets, we showed that regardless of batch information, SCORE outperforms existing methods in accuracy, robustness, scalability and data integration.

Description: Quantifying cell proportions, especially for rare cell types in some scenarios, is of great value in tracking signals associated with certain phenotypes or diseases. Although some methods have been proposed to infer cell proportions from multicomponent bulk data, they are substantially less effective for estimating the proportions of rare cell types which are highly sensitive to feature outliers and collinearity. Here we proposed a new deconvolution algorithm named ARIC to estimate cell type proportions from gene expression or DNA methylation data. ARIC employs a novel two-step marker selection strategy, including collinear feature elimination based on the component-wise condition number and adaptive removal of outlier markers. This strategy can systematically obtain effective markers for weighted $upsilon$-support vector regression to ensure a robust and precise rare proportion prediction. We showed that ARIC can accurately estimate fractions in both DNA methylation and gene expression data from different experiments. We further applied ARIC to the survival prediction of ovarian cancer and the condition monitoring of chronic kidney disease, and the results demonstrate the high accuracy and robustness as well as clinical potentials of ARIC. Taken together, ARIC is a promising tool to solve the deconvolution problem of bulk data where rare components are of vital importance.

Description: For an increasing number of preclinical samples, both detailed molecular profiles and their responses to various drugs are becoming available. Efforts to understand, and predict, drug responses in a data-driven manner have led to a proliferation of machine learning (ML) methods, with the longer term ambition of predicting clinical drug responses. Here, we provide a uniquely wide and deep systematic review of the rapidly evolving literature on monotherapy drug response prediction, with a systematic characterization and classification that comprises more than 70 ML methods in 13 subclasses, their input and output data types, modes of evaluation, and code and software availability. ML experts are provided with a fundamental understanding of the biological problem, and how ML methods are configured for it. Biologists and biomedical researchers are introduced to the basic principles of applicable ML methods, and their application to the problem of drug response prediction. We also provide systematic overviews of commonly used data sources used for training and evaluation methods.

Description: The biological functions of DNA and RNA generally depend on their interactions with other molecules, such as small ligands, proteins and nucleic acids. However, our knowledge of the nucleic acid binding sites for different interaction partners is very limited, and identification of these critical binding regions is not a trivial work. Herein, we performed a comprehensive comparison between binding and nonbinding sites and among different categories of binding sites in these two nucleic acid classes. From the structural perspective, RNA may interact with ligands through forming binding pockets and contact proteins and nucleic acids using protruding surfaces, while DNA may adopt regions closer to the middle of the chain to make contacts with other molecules. Based on structural information, we established a feature-based ensemble learning classifier to identify the binding sites by fully using the interplay among different machine learning algorithms, feature spaces and sample spaces. Meanwhile, we designed a template-based classifier by exploiting structural conservation. The complementarity between the two classifiers motivated us to build an integrative framework for improving prediction performance. Moreover, we utilized a post-processing procedure based on the random walk algorithm to further correct the integrative predictions. Our unified prediction framework yielded promising results for different binding sites and outperformed existing methods.

Description: Long-read sequencing technology enables significant progress in de novo genome assembly. However, the high error rate and the wide error distribution of raw reads result in a large number of errors in the assembly. Polishing is a procedure to fix errors in the draft assembly and improve the reliability of genomic analysis. However, existing methods treat all the regions of the assembly equally while there are fundamental differences between the error distributions of these regions. How to achieve very high accuracy in genome assembly is still a challenging problem. Motivated by the uneven errors in different regions of the assembly, we propose a novel polishing workflow named BlockPolish. In this method, we divide contigs into blocks with low complexity and high complexity according to statistics of aligned nucleotide bases. Multiple sequence alignment is applied to realign raw reads in complex blocks and optimize the alignment result. Due to the different distributions of error rates in trivial and complex blocks, two multitask bidirectional Long short-term memory (LSTM) networks are proposed to predict the consensus sequences. In the whole-genome assemblies of NA12878 assembled by Wtdbg2 and Flye using Nanopore data, BlockPolish has a higher polishing accuracy than other state-of-the-arts including Racon, Medaka and MarginPolish & HELEN. In all assemblies, errors are predominantly indels and BlockPolish has a good performance in correcting them. In addition to the Nanopore assemblies, we further demonstrate that BlockPolish can also reduce the errors in the PacBio assemblies. The source code of BlockPolish is freely available on Github (https://github.com/huangnengCSU/BlockPolish).

Description: Ordinary differential equation models are nowadays widely used for the mechanistic description of biological processes and their temporal evolution. These models typically have many unknown and nonmeasurable parameters, which have to be determined by fitting the model to experimental data. In order to perform this task, known as parameter estimation or model calibration, the modeller faces challenges such as poor parameter identifiability, lack of sufficiently informative experimental data and the existence of local minima in the objective function landscape. These issues tend to worsen with larger model sizes, increasing the computational complexity and the number of unknown parameters. An incorrectly calibrated model is problematic because it may result in inaccurate predictions and misleading conclusions. For nonexpert users, there are a large number of potential pitfalls. Here, we provide a protocol that guides the user through all the steps involved in the calibration of dynamic models. We illustrate the methodology with two models and provide all the code required to reproduce the results and perform the same analysis on new models. Our protocol provides practitioners and researchers in biological modelling with a one-stop guide that is at the same time compact and sufficiently comprehensive to cover all aspects of the problem.

Description: Complex computation and approximate solution hinder the application of generalized linear mixed models (GLMM) into genome-wide association studies. We extended GRAMMAR to handle binary diseases by considering genomic breeding values (GBVs) estimated in advance as a known predictor in genomic logit regression, and then reduced polygenic effects by regulating downward genomic heritability to control false negative errors produced in the association tests. Using simulations and case analyses, we showed in optimizing GRAMMAR, polygenic effects and genomic controls could be evaluated using the fewer sampling markers, which extremely simplified GLMM-based association analysis in large-scale data. Further, joint association analysis for quantitative trait nucleotide (QTN) candidates chosen by multiple testing offered significant improved statistical power to detect QTNs over existing methods.

Description: Identifying disease-related microRNAs (miRNAs) assists the understanding of disease pathogenesis. Existing research methods integrate multiple kinds of data related to miRNAs and diseases to infer candidate disease-related miRNAs. The attributes of miRNA nodes including their family and cluster belonging information, however, have not been deeply integrated. Besides, the learning of neighbor topology representation of a pair of miRNA and disease is a challenging issue. We present a disease-related miRNA prediction method by encoding and integrating multiple representations of miRNA and disease nodes learnt from the generative and adversarial perspective. We firstly construct a bilayer heterogeneous network of miRNA and disease nodes, and it contains multiple types of connections among these nodes, which reflect neighbor topology of miRNA–disease pairs, and the attributes of miRNA nodes, especially miRNA-related families and clusters. To learn enhanced pairwise neighbor topology, we propose a generative and adversarial model with a convolutional autoencoder-based generator to encode the low-dimensional topological representation of the miRNA–disease pair and multi-layer convolutional neural network-based discriminator to discriminate between the true and false neighbor topology embeddings. Besides, we design a novel feature category-level attention mechanism to learn the various importance of different features for final adaptive fusion and prediction. Comparison results with five miRNA–disease association methods demonstrated the superior performance of our model and technical contributions in terms of area under the receiver operating characteristic curve and area under the precision-recall curve. The results of recall rates confirmed that our model can find more actual miRNA–disease associations among top-ranked candidates. Case studies on three cancers further proved the ability to detect potential candidate miRNAs.

Description: Development of interactive web applications to deposit, visualize and analyze biological datasets is a major subject of bioinformatics. R is a programming language for data science, which is also one of the most popular languages used in biological data analysis and bioinformatics. However, building interactive web applications was a great challenge for R users before the Shiny package was developed by the RStudio company in 2012. By compiling R code into HTML, CSS and JavaScript code, Shiny has made it incredibly easy to build web applications for the large R community in bioinformatics and for even non-programmers. Over 470 biological web applications have been developed with R/Shiny up to now. To further promote the utilization of R/Shiny, we reviewed the development of biological web applications with R/Shiny, including eminent biological web applications built with R/Shiny, basic steps to build an R/Shiny application, commonly used R packages to build the interface and server of R/Shiny applications, deployment of R/Shiny applications in the cloud and online resources for R/Shiny.

Description: Understanding the impact of non-coding sequence variants on complex diseases is an essential problem. We present a novel ensemble learning framework—CASAVA, to predict genomic loci in terms of disease category-specific risk. Using disease-associated variants identified by GWAS as training data, and diverse sequencing-based genomics and epigenomics profiles as features, CASAVA provides risk prediction of 24 major categories of diseases throughout the human genome. Our studies showed that CASAVA scores at a genomic locus provide a reasonable prediction of the disease-specific and disease category-specific risk prediction for non-coding variants located within the locus. Taking MHC2TA and immune system diseases as an example, we demonstrate the potential of CASAVA in revealing variant-disease associations. A website (http://zhanglabtools.org/CASAVA) has been built to facilitate easily access to CASAVA scores.

Description: Cancer is thought to be caused by the accumulation of driver genetic mutations. Therefore, identifying cancer driver genes plays a crucial role in understanding the molecular mechanism of cancer and developing precision therapies and biomarkers. In this work, we propose a Multi-Task learning method, called MTGCN, based on the Graph Convolutional Network to identify cancer driver genes. First, we augment gene features by introducing their features on the protein-protein interaction (PPI) network. After that, the multi-task learning framework propagates and aggregates nodes and graph features from input to next layer to learn node embedding features, simultaneously optimizing the node prediction task and the link prediction task. Finally, we use a Bayesian task weight learner to balance the two tasks automatically. The outputs of MTGCN assign each gene a probability of being a cancer driver gene. Our method and the other four existing methods are applied to predict cancer drivers for pan-cancer and some single cancer types. The experimental results show that our model shows outstanding performance compared with the state-of-the-art methods in terms of the area under the Receiver Operating Characteristic (ROC) curves and the area under the precision-recall curves. The MTGCN is freely available via https://github.com/weiba/MTGCN.

Description: Protein post-translational modifications (PTM) play vital roles in cellular regulation, modulating functions by driving changes in protein structure and dynamics. Exploring comprehensively the influence of PTM on conformational dynamics can facilitate the understanding of the related biological function and molecular mechanism. Currently, a series of excellent computation tools have been designed to analyze the time-dependent structural properties of proteins. However, the protocol aimed to explore conformational dynamics of post-translational modified protein is still a blank. To fill this gap, we present PTMdyna to visually predict the conformational dynamics differences between unmodified and modified proteins, thus indicating the influence of specific PTM. PTMdyna exhibits an AUC of 0.884 tested on 220 protein–protein complex structures. The case of heterochromatin protein 1α complexed with lysine 9-methylated histone H3, which is critical for genomic stability and cell differentiation, was used to demonstrate its applicability. PTMdyna provides a reliable platform to predict the influence of PTM on protein dynamics, making it easier to interpret PTM functionality at the structure level. The web server is freely available at http://ccbportal.com/PTMdyna.

Description: MicroRNAs (miRNAs) play crucial roles in human disease and can be targeted by small molecule (SM) drugs according to numerous studies, which shows that identifying SM–miRNA associations in human disease is important for drug development and disease treatment. We proposed the method of Ensemble of Kernel Ridge Regression-based Small Molecule–MiRNA Association prediction (EKRRSMMA) to uncover potential SM–miRNA associations by combing feature dimensionality reduction and ensemble learning. First, we constructed different feature subsets for both SMs and miRNAs. Then, we trained homogeneous base learners based on distinct feature subsets and took the average of scores obtained from these base learners as SM–miRNA association score. In EKRRSMMA, feature dimensionality reduction technology was employed in the process of construction of feature subsets to reduce the influence of noisy data. Besides, the base learner, namely KRR_avg, was the combination of two classifiers constructed under SM space and miRNA space, which could make full use of the information of SM and miRNA. To assess the prediction performance of EKRRSMMA, we conducted Leave-One-Out Cross-Validation (LOOCV), SM-fixed local LOOCV, miRNA-fixed local LOOCV and 5-fold CV based on two datasets. For Dataset 1 (Dataset 2), EKRRSMMA got the Area Under receiver operating characteristic Curves (AUCs) of 0.9793 (0.8871), 0.8071 (0.7705), 0.9732 (0.8586) and 0.9767 ± 0.0014 (0.8560 ± 0.0027). Besides, we conducted four case studies. As a result, 32 (5-Fluorouracil), 19 (17β-Estradiol), 26 (5-Aza-2′-deoxycytidine) and 11 (cyclophosphamide) out of top 50 predicted potentially associated miRNAs were confirmed by database or experimental literature. Above evaluation results demonstrated that EKRRSMMA is reliable for predicting SM–miRNA associations.

Description: The employment of doubled-haploid (DH) technology in maize has vastly accelerated the efficiency of developing inbred lines. The selection of superior lines has to rely on genotypes with genomic selection (GS) model, rather than phenotypes due to the high expense of field phenotyping. In this work, we implemented ‘genome optimization via virtual simulation (GOVS)’ using the genotype and phenotype data of 1404 maize lines and their F1 progeny. GOVS simulates a virtual genome encompassing the most abundant ‘optimal genotypes’ or ‘advantageous alleles’ in a genetic pool. Such a virtually optimized genome, although can never be developed in reality, may help plot the optimal route to direct breeding decisions. GOVS assists in the selection of superior lines based on the genomic fragments that a line contributes to the simulated genome. The assumption is that the more fragments of optimal genotypes a line contributes to the assembly, the higher the likelihood of the line favored in the F1 phenotype, e.g. grain yield. Compared to traditional GS method, GOVS-assisted selection may avoid using an arbitrary threshold for the predicted F1 yield to assist selection. Additionally, the selected lines contributed complementary sets of advantageous alleles to the virtual genome. This feature facilitates plotting the optimal route for DH production, whereby the fewest lines and F1 combinations are needed to pyramid a maximum number of advantageous alleles in the new DH lines. In summary, incorporation of DH production, GS and genome optimization will ultimately improve genomically designed breeding in maize. Short abstract: Doubled-haploid (DH) technology has been widely applied in maize breeding industry, as it greatly shortens the period of developing homozygous inbred lines via bypassing several rounds of self-crossing. The current challenge is how to efficiently screen the large volume of inbred lines based on genotypes. We present the toolbox of genome optimization via virtual simulation (GOVS), which complements the traditional genomic selection model. GOVS simulates a virtual genome encompassing the most abundant ‘optimal genotypes’ in a breeding population, and then assists in selection of superior lines based on the genomic fragments that a line contributes to the simulated genome. Availability of GOVS (https://govs-pack.github.io/) to the public may ultimately facilitate genomically designed breeding in maize.

Description: Circular RNAs (circRNAs) are a category of novelty discovered competing endogenous non-coding RNAs that have been proved to implicate many human complex diseases. A large number of circRNAs have been confirmed to be involved in cancer progression and are expected to become promising biomarkers for tumor diagnosis and targeted therapy. Deciphering the underlying relationships between circRNAs and diseases may provide new insights for us to understand the pathogenesis of complex diseases and further characterize the biological functions of circRNAs. As traditional experimental methods are usually time-consuming and laborious, computational models have made significant progress in systematically exploring potential circRNA-disease associations, which not only creates new opportunities for investigating pathogenic mechanisms at the level of circRNAs, but also helps to significantly improve the efficiency of clinical trials. In this review, we first summarize the functions and characteristics of circRNAs and introduce some representative circRNAs related to tumorigenesis. Then, we mainly investigate the available databases and tools dedicated to circRNA and disease studies. Next, we present a comprehensive review of computational methods for predicting circRNA-disease associations and classify them into five categories, including network propagating-based, path-based, matrix factorization-based, deep learning-based and other machine learning methods. Finally, we further discuss the challenges and future researches in this field.

Description: Single-cell RNA sequencing (scRNA-Seq) is an emerging strategy for characterizing immune cell populations. Compared to flow or mass cytometry, scRNA-Seq could potentially identify cell types and activation states that lack precise cell surface markers. However, scRNA-Seq is currently limited due to the need to manually classify each immune cell from its transcriptional profile. While recently developed algorithms accurately annotate coarse cell types (e.g. T cells versus macrophages), making fine distinctions (e.g. CD8+ effector memory T cells) remains a difficult challenge. To address this, we developed a machine learning classifier called ImmClassifier that leverages a hierarchical ontology of cell type. We demonstrate that its predictions are highly concordant with flow-based markers from CITE-seq and outperforms other tools (+15% recall, +14% precision) in distinguishing fine-grained cell types with comparable performance on coarse ones. Thus, ImmClassifier can be used to explore more deeply the heterogeneity of the immune system in scRNA-Seq experiments.

Description: Drug–drug interactions (DDIs) are interactions with adverse effects on the body, manifested when two or more incompatible drugs are taken together. They can be caused by the chemical compositions of the drugs involved. We introduce gated message passing neural network (GMPNN), a message passing neural network which learns chemical substructures with different sizes and shapes from the molecular graph representations of drugs for DDI prediction between a pair of drugs. In GMPNN, edges are considered as gates which control the flow of message passing, and therefore delimiting the substructures in a learnable way. The final DDI prediction between a drug pair is based on the interactions between pairs of their (learned) substructures, each pair weighted by a relevance score to the final DDI prediction output. Our proposed method GMPNN-CS (i.e. GMPNN + prediction module) is evaluated on two real-world datasets, with competitive results on one, and improved performance on the other compared with previous methods. Source code is freely available at https://github.com/kanz76/GMPNN-CS.

Description: We present a concatenated deep-learning multiple neural network system for the analysis of single-molecule trajectories. We apply this machine learning-based analysis to characterize the translational diffusion of the nicotinic acetylcholine receptor at the plasma membrane, experimentally interrogated using superresolution optical microscopy. The receptor protein displays a heterogeneous diffusion behavior that goes beyond the ensemble level, with individual trajectories exhibiting more than one diffusive state, requiring the optimization of the neural networks through a hyperparameter analysis for different numbers of steps and durations, especially for short trajectories (