Dosage-sensitive molecular mechanisms are associated with the tissue-specificity of traits and diseases

Tissue-aware interpretation of genetic variants advances the etiology of rare diseases

Pathogenic variants underlying Mendelian diseases often disrupt the normal physiology of a few tissues and organs. However, variant effect prediction tools that aim to identify pathogenic variants are typically oblivious to tissue contexts. Here we report a machine-learning framework, denoted "Tissue Risk Assessment of Causality by Expression for variants" (TRACEvar, https://netbio.bgu.ac.il/TRACEvar/ ), that offers two advancements. First, TRACEvar predicts pathogenic variants that disrupt the normal physiology of specific tissues. This was achieved by creating 14 tissue-specific models that were trained on over 14,000 variants and combined 84 attributes of genetic variants with 495 attributes derived from tissue omics. TRACEvar outperformed 10 well-established and tissue-oblivious variant effect prediction tools. Second, the resulting models are interpretable, thereby illuminating variants' mode of action. Application of TRACEvar to variants of 52 rare-disease patients highlighted pathogenicity mechanisms and relevant disease processes. Lastly, the interpretation of all tissue models revealed that top-ranking determinants of pathogenicity included attributes of disease-affected tissues, particularly cellular process activities. Collectively, these results show that tissue contexts and interpretable machine-learning models can greatly enhance the etiology of rare diseases.

Network-based modelling reveals cell-type enriched patterns of non-coding RNA regulation during human skeletal muscle remodelling

A majority of human genes produce non-protein-coding RNA (ncRNA), and some have roles in development and disease. Neither ncRNA nor human skeletal muscle is ideally studied using short-read sequencing, so we used a customised RNA pipeline and network modelling to study cell-type specific ncRNA responses during muscle growth at scale. We completed five human resistance-training studies (n=144 subjects), identifying 61% who successfully accrued muscle-mass. We produced 288 transcriptome-wide profiles and found 110 ncRNAs linked to muscle growth in vivo, while a transcriptome-driven network model demonstrated interactions via a number of discrete functional pathways and single-cell types. This analysis included established hypertrophy-related ncRNAs, including CYTOR - which was leukocyte-associated (FDR = 4.9 ×10-7). Novel hypertrophy-linked ncRNAs included PPP1CB-DT (myofibril assembly genes, FDR = 8.15 × 10-8), and EEF1A1P24 and TMSB4XP8 (vascular remodelling and angiogenesis genes, FDR = 2.77 × 10-5). We also discovered that hypertrophy lncRNA MYREM shows a specific myonuclear expression pattern in vivo. Our multi-layered analyses established that single-cell-associated ncRNA are identifiable from bulk muscle transcriptomic data and that hypertrophy-linked ncRNA genes mediate their association with muscle growth via multiple cell types and a set of interacting pathways.

Network-based modelling reveals cell-type enriched patterns of non-coding RNA regulation during human skeletal muscle remodelling

A majority of human genes produce non-protein-coding RNA (ncRNA), and some have roles in development and disease. Neither ncRNA nor human skeletal muscle is ideally studied using short-read sequencing, so we used a customized RNA pipeline and network modelling to study cell-type specific ncRNA responses during muscle growth at scale. We completed five human resistance-training studies (n = 144 subjects), identifying 61% who successfully accrued muscle-mass. We produced 288 transcriptome-wide profiles and found 110 ncRNAs linked to muscle growth in vivo, while a transcriptome-driven network model demonstrated interactions via a number of discrete functional pathways and single-cell types. This analysis included established hypertrophy-related ncRNAs, including CYTOR-which was leukocyte-associated (false discovery rate [FDR] = 4.9 × 10-7). Novel hypertrophy-linked ncRNAs included PPP1CB-DT (myofibril assembly genes, FDR = 8.15 × 10-8), and EEF1A1P24 and TMSB4XP8 (vascular remodelling and angiogenesis genes, FDR = 2.77 × 10-5). We also discovered that hypertrophy lncRNA MYREM shows a specific myonuclear expression pattern in vivo. Our multi-layered analyses established that single-cell-associated ncRNA are identifiable from bulk muscle transcriptomic data and that hypertrophy-linked ncRNA genes mediate their association with muscle growth via multiple cell types and a set of interacting pathways.

Machine-learning analysis reveals an important role for negative selection in shaping cancer aneuploidy landscapes

BACKGROUND

Aneuploidy, an abnormal number of chromosomes within a cell, is a hallmark of cancer. Patterns of aneuploidy differ across cancers, yet are similar in cancers affecting closely related tissues. The selection pressures underlying aneuploidy patterns are not fully understood, hindering our understanding of cancer development and progression.

RESULTS

Here, we apply interpretable machine learning methods to study tissue-selective aneuploidy patterns. We define 20 types of features corresponding to genomic attributes of chromosome-arms, normal tissues, primary tumors, and cancer cell lines (CCLs), and use them to model gains and losses of chromosome arms in 24 cancer types. To reveal the factors that shape the tissue-specific cancer aneuploidy landscapes, we interpret the machine learning models by estimating the relative contribution of each feature to the models. While confirming known drivers of positive selection, our quantitative analysis highlights the importance of negative selection for shaping aneuploidy landscapes. This is exemplified by tumor suppressor gene density being a better predictor of gain patterns than oncogene density, and vice versa for loss patterns. We also identify the importance of tissue-selective features and demonstrate them experimentally, revealing KLF5 as an important driver for chr13q gain in colon cancer. Further supporting an important role for negative selection in shaping the aneuploidy landscapes, we find compensation by paralogs to be among the top predictors of chromosome arm loss prevalence and demonstrate this relationship for one paralog interaction. Similar factors shape aneuploidy patterns in human CCLs, demonstrating their relevance for aneuploidy research.

CONCLUSIONS

Our quantitative, interpretable machine learning models improve the understanding of the genomic properties that shape cancer aneuploidy landscapes.

Predicting molecular mechanisms of hereditary diseases by using their tissue-selective manifestation

How do aberrations in widely expressed genes lead to tissue-selective hereditary diseases? Previous attempts to answer this question were limited to testing a few candidate mechanisms. To answer this question at a larger scale, we developed "Tissue Risk Assessment of Causality by Expression" (TRACE), a machine learning approach to predict genes that underlie tissue-selective diseases and selectivity-related features. TRACE utilized 4,744 biologically interpretable tissue-specific gene features that were inferred from heterogeneous omics datasets. Application of TRACE to 1,031 disease genes uncovered known and novel selectivity-related features, the most common of which was previously overlooked. Next, we created a catalog of tissue-associated risks for 18,927 protein-coding genes (https://netbio.bgu.ac.il/trace/). As proof-of-concept, we prioritized candidate disease genes identified in 48 rare-disease patients. TRACE ranked the verified disease gene among the patient's candidate genes significantly better than gene prioritization methods that rank by gene constraint or tissue expression. Thus, tissue selectivity combined with machine learning enhances genetic and clinical understanding of hereditary diseases.

Transcriptomics for Clinical and Experimental Biology Research: Hang on a Seq

Sequencing the human genome empowers translational medicine, facilitating transcriptome-wide molecular diagnosis, pathway biology, and drug repositioning. Initially, microarrays are used to study the bulk transcriptome; but now short-read RNA sequencing (RNA-seq) predominates. Positioned as a superior technology, that makes the discovery of novel transcripts routine, most RNA-seq analyses are in fact modeled on the known transcriptome. Limitations of the RNA-seq methodology have emerged, while the design of, and the analysis strategies applied to, arrays have matured. An equitable comparison between these technologies is provided, highlighting advantages that modern arrays hold over RNA-seq. Array protocols more accurately quantify constitutively expressed protein coding genes across tissue replicates, and are more reliable for studying lower expressed genes. Arrays reveal long noncoding RNAs (lncRNA) are neither sparsely nor lower expressed than protein coding genes. Heterogeneous coverage of constitutively expressed genes observed with RNA-seq, undermines the validity and reproducibility of pathway analyses. The factors driving these observations, many of which are relevant to long-read or single-cell sequencing are discussed. As proposed herein, a reappreciation of bulk transcriptomic methods is required, including wider use of the modern high-density array data-to urgently revise existing anatomical RNA reference atlases and assist with more accurate study of lncRNAs.

Paralog Explorer: A resource for mining information about paralogs in common research organisms

Paralogs are genes which arose via gene duplication, and when such paralogs retain overlapping or redundant function, this poses a challenge to functional genetics research. Recent technological advancements have made it possible to systematically probe gene function for redundant genes using dual or multiplex gene perturbation, and there is a need for a simple bioinformatic tool to identify putative paralogs of a gene(s) of interest. We have developed Paralog Explorer (https://www.flyrnai.org/tools/paralogs/), an online resource that allows researchers to quickly and accurately identify candidate paralogous genes in the genomes of the model organisms D. melanogaster, C. elegans, D. rerio, M. musculus, and H. sapiens. Paralog Explorer deploys an effective between-species ortholog prediction software, DIOPT, to analyze within-species paralogs. Paralog Explorer allows users to identify candidate paralogs, and to navigate relevant databases regarding gene co-expression, protein-protein and genetic interaction, as well as gene ontology and phenotype annotations. Altogether, this tool extends the value of current ortholog prediction resources by providing sophisticated features useful for identification and study of paralogous genes.

The Organ-Disease Annotations (ODiseA) database of hereditary diseases and inflicted tissues

Hereditary diseases tend to manifest clinically in few selected tissues. Knowledge of those tissues is important for better understanding of disease mechanisms, which often remain elusive. However, information on the tissues inflicted by each disease is not easily obtainable. Well-established resources, such as the Online Mendelian Inheritance in Man (OMIM) database and Human Phenotype Ontology (HPO), report on a spectrum of disease manifestations, yet do not highlight the main inflicted tissues. The Organ-Disease Annotations (ODiseA) database contains 4,357 thoroughly-curated annotations for 2,181 hereditary diseases and 45 inflicted tissues. Additionally, ODiseA reports 692 annotations of 635 diseases and the pathogenic tissues where they emerge. ODiseA can be queried by disease, disease gene, or inflicted tissue. Owing to its expansive, high-quality annotations, ODiseA serves as a valuable and unique tool for biomedical and computational researchers studying genotype-phenotype relationships of hereditary diseases. ODiseA is available at https://netbio.bgu.ac.il/odisea.

The differential activity of biological processes in tissues and cell subsets can illuminate disease-related processes and cell-type identities

MOTIVATION

The distinct functionalities of human tissues and cell types underlie complex phenotype-genotype relationships, yet often remain elusive. Harnessing the multitude of bulk and single-cell human transcriptomes while focusing on processes can help reveal these distinct functionalities.

RESULTS

The Tissue-Process Activity (TiPA) method aims to identify processes that are preferentially active or under-expressed in specific contexts, by comparing the expression levels of process genes between contexts. We tested TiPA on 1579 tissue-specific processes and bulk tissue transcriptomes, finding that it performed better than another method. Next, we used TiPA to ask whether the activity of certain processes could underlie the tissue-specific manifestation of 1233 hereditary diseases. We found that 21% of the disease-causing genes indeed participated in such processes, thereby illuminating their genotype-phenotype relationships. Lastly, we applied TiPA to single-cell transcriptomes of 108 human cell types, revealing that process activities often match cell-type identities and can thus aid annotation efforts. Hence, differential activity of processes can highlight the distinct functionality of tissues and cells in a robust and meaningful manner.

AVAILABILITY AND IMPLEMENTATION

TiPA code is available in GitHub (https://github.com/moranshar/TiPA). In addition, all data are available as part of the Supplementary Material.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.