Challenges in Bioinformatics Workflows for Processing Microbiome Omics Data at Scale

The nascent field of microbiome science is transitioning from a descriptive approach of cataloging taxa and functions present in an environment to applying multi-omics methods to investigate microbiome dynamics and function. A large number of new tools and algorithms have been designed and used for very specific purposes on samples collected by individual investigators or groups. While these developments have been quite instructive, the ability to compare microbiome data generated by many groups of researchers is impeded by the lack of standardized application of bioinformatics methods. Additionally, there are few examples of broad bioinformatics workflows that can process metagenome, metatranscriptome, metaproteome and metabolomic data at scale, and no central hub that allows processing, or provides varied omics data that are findable, accessible, interoperable and reusable (FAIR). Here, we review some of the challenges that exist in analyzing omics data within the microbiome research sphere, and provide context on how the National Microbiome Data Collaborative has adopted a standardized and open access approach to address such challenges.

Aristotle: stratified causal discovery for omics data

Background There has been a simultaneous increase in demand and accessibility across genomics, transcriptomics, proteomics and metabolomics data, known as omics data. This has encouraged widespread application of omics data in life sciences, from personalized medicine to the discovery of underlying pathophysiology of diseases. Causal analysis of omics data may provide important insight into the underlying biological mechanisms. Existing causal analysis methods yield promising results when identifying potential general causes of an observed outcome based on omics data. However, they may fail to discover the causes specific to a particular stratum of individuals and missing from others. Methods To fill this gap, we introduce the problem of stratified causal discovery and propose a method, Aristotle, for solving it. Aristotle addresses the two challenges intrinsic to omics data: high dimensionality and hidden stratification. It employs existing biological knowledge and a state-of-the-art patient stratification method to tackle the above challenges and applies a quasi-experimental design method to each stratum to find stratum-specific potential causes. Results Evaluation based on synthetic data shows better performance for Aristotle in discovering true causes under different conditions compared to existing causal discovery methods. Experiments on a real dataset on Anthracycline Cardiotoxicity indicate that Aristotle’s predictions are consistent with the existing literature. Moreover, Aristotle makes additional predictions that suggest further investigations.

Probabilistic modeling of ambient noise in single-cell omics data

Droplet-based single-cell omics, including single-cell RNA sequencing (scRNAseq), single cell CRISPR perturbations (e.g., CROP-seq) and single-cell protein and transcriptomic profiling (e.g., CITE-seq) hold great promise for comprehensive cell profiling and genetic screening at the single cell resolution, yet these technologies suffer from substantial noise, among which ambient signals present in the cell suspension may be the predominant source. Current efforts to address this issue are highly specific to a certain technology, while a universal model to describe the noise across these technologies may reveal this common source thereby improving the denoising accuracy. To this end, we explicitly examined these unexpected signals and observed a predictable pattern in multiple datasets across different technologies. Based on the finding, we developed single cell Ambient Remover (scAR) which uses probabilistic deep learning to deconvolute the observed signals into native and ambient composition. scAR provides an efficient and universal solution to count denoising for multiple types of single-cell omics data, including single cell CRISPR screens, CITE-seq and scRNAseq. It will facilitate the application of single-cell omics technologies.

InterTADs: integration of multi-omics data on topologically associated domains, application to chronic lymphocytic leukemia

ABSTRACT The integration of multi-omics data can greatly facilitate the advancement of research in Life Sciences by highlighting new interactions. However, there is currently no widespread procedure for meaningful multi-omics data integration. Here, we present a robust framework, called InterTADs, for integrating multi-omics data derived from the same sample, and considering the chromatin configuration of the genome, i.e. the topologically associating domains (TADs). Following the integration process, statistical analysis highlights the differences between the groups of interest (normal versus cancer cells) relating to (i) independent and (ii) integrated events through TADs. Finally, enrichment analysis using KEGG database, Gene Ontology and transcription factor binding sites and visualization approaches are available. We applied InterTADs to multi-omics datasets from 135 patients with chronic lymphocytic leukemia (CLL) and found that the integration through TADs resulted in a dramatic reduction of heterogeneity compared to individual events. Significant differences for individual events and on TADs level were identified between patients differing in the somatic hypermutation status of the clonotypic immunoglobulin genes, the core biological stratifier in CLL, attesting to the biomedical relevance of InterTADs. In conclusion, our approach suggests a new perspective towards analyzing multi-omics data, by offering reasonable execution time, biological benchmarking and potentially contributing to pattern discovery through TADs.

ChromoMap: an R package for interactive visualization of multi-omics data and annotation of chromosomes

Background The recent advancements in high-throughput sequencing have resulted in the availability of annotated genomes, as well as of multi-omics data for many living organisms. This has increased the need for graphic tools that allow the concurrent visualization of genomes and feature-associated multi-omics data on single publication-ready plots. Results We present chromoMap, an R package, developed for the construction of interactive visualizations of chromosomes/chromosomal regions, mapping of any chromosomal feature with known coordinates (i.e., protein coding genes, transposable elements, non-coding RNAs, microsatellites, etc.), and chromosomal regional characteristics (i.e. genomic feature density, gene expression, DNA methylation, chromatin modifications, etc.) of organisms with a genome assembly. ChromoMap can also integrate multi-omics data (genomics, transcriptomics and epigenomics) in relation to their occurrence across chromosomes. ChromoMap takes tab-delimited files (BED like) or alternatively R objects to specify the genomic co-ordinates of the chromosomes and elements to annotate. Rendered chromosomes are composed of continuous windows of a given range, which, on hover, display detailed information about the elements annotated within that range. By adjusting parameters of a single function, users can generate a variety of plots that can either be saved as static image or as HTML documents. Conclusions ChromoMap’s flexibility allows for concurrent visualization of genomic data in each strand of a given chromosome, or of more than one homologous chromosome; allowing the comparison of multi-omic data between genotypes (e.g. species, varieties, etc.) or between homologous chromosomes of phased diploid/polyploid genomes. chromoMap is an extensive tool that can be potentially used in various bioinformatics analysis pipelines for genomic visualization of multi-omics data.

Artificial intelligence framework identifies candidate targets for drug repurposing in Alzheimer’s disease

Background Genome-wide association studies (GWAS) have identified numerous susceptibility loci for Alzheimer’s disease (AD). However, utilizing GWAS and multi-omics data to identify high-confidence AD risk genes (ARGs) and druggable targets that can guide development of new therapeutics for patients suffering from AD has heretofore not been successful. Methods To address this critical problem in the field, we have developed a network-based artificial intelligence framework that is capable of integrating multi-omics data along with human protein–protein interactome networks to accurately infer accurate drug targets impacted by GWAS-identified variants to identify new therapeutics. When applied to AD, this approach integrates GWAS findings, multi-omics data from brain samples of AD patients and AD transgenic animal models, drug-target networks, and the human protein–protein interactome, along with large-scale patient database validation and in vitro mechanistic observations in human microglia cells. Results Through this approach, we identified 103 ARGs validated by various levels of pathobiological evidence in AD. Via network-based prediction and population-based validation, we then showed that three drugs (pioglitazone, febuxostat, and atenolol) are significantly associated with decreased risk of AD compared with matched control populations. Pioglitazone usage is significantly associated with decreased risk of AD (hazard ratio (HR) = 0.916, 95% confidence interval [CI] 0.861–0.974, P = 0.005) in a retrospective case-control validation. Pioglitazone is a peroxisome proliferator-activated receptor (PPAR) agonist used to treat type 2 diabetes, and propensity score matching cohort studies confirmed its association with reduced risk of AD in comparison to glipizide (HR = 0.921, 95% CI 0.862–0.984, P = 0.0159), an insulin secretagogue that is also used to treat type 2 diabetes. In vitro experiments showed that pioglitazone downregulated glycogen synthase kinase 3 beta (GSK3β) and cyclin-dependent kinase (CDK5) in human microglia cells, supporting a possible mechanism-of-action for its beneficial effect in AD. Conclusions In summary, we present an integrated, network-based artificial intelligence methodology to rapidly translate GWAS findings and multi-omics data to genotype-informed therapeutic discovery in AD.

Blocking hepatocarcinogenesis by a cytochrome P450 family member with female-preferential expression

ObjectsThe incidence of hepatocellular carcinoma (HCC) shows an obvious male dominance in rodents and humans. We aimed to identify the key autosomal liver-specific sex-related genes and investigate their roles in hepatocarcinogenesis.DesignTwo HCC cohorts (n=551) with available transcriptome and metabolome data were used. Class comparisons of omics data and ingenuity pathway analysis were performed to explore sex-related molecules and their associated functions. Functional assays were employed to investigate roles of the key candidates, including cellular assays, molecular assays and multiple orthotopic HCC mouse models.ResultsA global comparison of multiple omics data revealed 861 sex-related molecules in non-tumour liver tissues between female and male HCC patients, which denoted a significant suppression of cancer-related diseases and functions in female liver than male. A member of cytochrome P450 family, CYP39A1, was one of the top liver-specific candidates with significantly higher levels in female vs male liver. In HCC tumours, CYP39A1 expression was dramatically reduced in over 90% HCC patients. Exogenous CYP39A1 significantly blocked tumour formation in both female and male mice and partially reduced the sex disparity of hepatocarcinogenesis. The HCC suppressor role of CYP39A1 did not rely on its known P450 enzyme activity but its C-terminal region, by which CYP39A1 impeded the transcriptional activation activity of c-Myc, leading to a significant inhibition of hepatocarcinogenesis.ConclusionsThe liver-specific CYP39A1 with female-preferential expression was a strong suppressor of HCC development. Strategies to up-regulate CYP39A1 might be promising methods for HCC treatment in both women and men in future.

Pan-Cancer DNA Methylation Analysis and Tumor Origin Identification of Carcinoma of Unknown Primary Site Based on Multi-Omics

The metastatic cancer of unknown primary (CUP) sites remains a leading cause of cancer death with few therapeutic options. The aberrant DNA methylation (DNAm) is the most important risk factor for cancer, which has certain tissue specificity. However, how DNAm alterations in tumors differ among the regulatory network of multi-omics remains largely unexplored. Therefore, there is room for improvement in our accuracy in the prediction of tumor origin sites and a need for better understanding of the underlying mechanisms. In our study, an integrative analysis based on multi-omics data and molecular regulatory network uncovered genome-wide methylation mechanism and identified 23 epi-driver genes. Apart from the promoter region, we also found that the aberrant methylation within the gene body or intergenic region was significantly associated with gene expression. Significant enrichment analysis of the epi-driver genes indicated that these genes were highly related to cellular mechanisms of tumorigenesis, including T-cell differentiation, cell proliferation, and signal transduction. Based on the ensemble algorithm, six CpG sites located in five epi-driver genes were selected to construct a tissue-specific classifier with a better accuracy (>95%) using TCGA datasets. In the independent datasets and the metastatic cancer datasets from GEO, the accuracy of distinguishing tumor subtypes or original sites was more than 90%, showing better robustness and stability. In summary, the integration analysis of large-scale omics data revealed complex regulation of DNAm across various cancer types and identified the epi-driver genes participating in tumorigenesis. Based on the aberrant methylation status located in epi-driver genes, a classifier that provided the highest accuracy in tracing back to the primary sites of metastatic cancer was established. Our study provides a comprehensive and multi-omics view of DNAm-associated changes across cancer types and has potential for clinical application.

Interpretable generative deep learning: an illustration with single cell gene expression data

Deep generative models can learn the underlying structure, such as pathways or gene programs, from omics data. We provide an introduction as well as an overview of such techniques, specifically illustrating their use with single-cell gene expression data. For example, the low dimensional latent representations offered by various approaches, such as variational auto-encoders, are useful to get a better understanding of the relations between observed gene expressions and experimental factors or phenotypes. Furthermore, by providing a generative model for the latent and observed variables, deep generative models can generate synthetic observations, which allow us to assess the uncertainty in the learned representations. While deep generative models are useful to learn the structure of high-dimensional omics data by efficiently capturing non-linear dependencies between genes, they are sometimes difficult to interpret due to their neural network building blocks. More precisely, to understand the relationship between learned latent variables and observed variables, e.g., gene transcript abundances and external phenotypes, is difficult. Therefore, we also illustrate current approaches that allow us to infer the relationship between learned latent variables and observed variables as well as external phenotypes. Thereby, we render deep learning approaches more interpretable. In an application with single-cell gene expression data, we demonstrate the utility of the discussed methods.