Proteins encoded in genomic regions associated with immune-mediated disease physically interact and suggest underlying biology

Molecular cascades and cell type-specific signatures in ASD revealed by single-cell genomics

Genomic profiling in postmortem brain from autistic individuals has consistently revealed convergent molecular changes. What drives these changes and how they relate to genetic susceptibility in this complex condition are not well understood. We performed deep single-nucleus RNA sequencing (snRNA-seq) to examine cell composition and transcriptomics, identifying dysregulation of cell type-specific gene regulatory networks (GRNs) in autism spectrum disorder (ASD), which we corroborated using single-nucleus assay for transposase-accessible chromatin with sequencing (snATAC-seq) and spatial transcriptomics. Transcriptomic changes were primarily cell type specific, involving multiple cell types, most prominently interhemispheric and callosal-projecting neurons, interneurons within superficial laminae, and distinct glial reactive states involving oligodendrocytes, microglia, and astrocytes. Autism-associated GRN drivers and their targets were enriched in rare and common genetic risk variants, connecting autism genetic susceptibility and cellular and circuit alterations in the human brain.

Nova proteins direct synaptic integration of somatostatin interneurons through activity-dependent alternative splicing

Somatostatin interneurons are the earliest born population of cortical inhibitory cells. They are crucial to support normal brain development and function; however, the mechanisms underlying their integration into nascent cortical circuitry are not well understood. In this study, we begin by demonstrating that the maturation of somatostatin interneurons in mouse somatosensory cortex is activity dependent. We then investigated the relationship between activity, alternative splicing, and synapse formation within this population. Specifically, we discovered that the Nova family of RNA-binding proteins are activity-dependent and are essential for the maturation of somatostatin interneurons, as well as their afferent and efferent connectivity. Within this population, Nova2 preferentially mediates the alternative splicing of genes required for axonal formation and synaptic function independently from its effect on gene expression. Hence, our work demonstrates that the Nova family of proteins through alternative splicing are centrally involved in coupling developmental neuronal activity to cortical circuit formation.

Comorbidity genetic risk and pathways impact SARS-CoV-2 infection outcomes

Understanding the genetic risk and mechanisms through which SARS-CoV-2 infection outcomes and comorbidities interact to impact acute and long-term sequelae is essential if we are to reduce the ongoing health burdens of the COVID-19 pandemic. Here we use a de novo protein diffusion network analysis coupled with tissue-specific gene regulatory networks, to examine putative mechanisms for associations between SARS-CoV-2 infection outcomes and comorbidities. Our approach identifies a shared genetic aetiology and molecular mechanisms for known and previously unknown comorbidities of SARS-CoV-2 infection outcomes. Additionally, genomic variants, genes and biological pathways that provide putative causal mechanisms connecting inherited risk factors for SARS-CoV-2 infection and coronary artery disease and Parkinson's disease are identified for the first time. Our findings provide an in depth understanding of genetic impacts on traits that collectively alter an individual's predisposition to acute and post-acute SARS-CoV-2 infection outcomes. The existence of complex inter-relationships between the comorbidities we identify raises the possibility of a much greater post-acute burden arising from SARS-CoV-2 infection if this genetic predisposition is realised.

A unifying statistical framework to discover disease genes from GWASs

Genome-wide association studies (GWASs) identify genomic loci associated with complex traits, but it remains a challenge to identify the genes affected by causal genetic variants in these loci. Attempts to solve this challenge are frustrated by a number of compounding problems. Here, we show how to combine solutions to these problems into a unified mathematical framework. From this synthesis, it becomes possible to compute the probability that each gene in the genome is affected by a causal variant, given a particular trait, without making assumptions about the relevant cell types or tissues. We validate each component of the framework individually and in combination. When applied to large GWASs of human disease, the resulting paradigm can rediscover the majority of well-known disease genes. Moreover, it establishes human genetics support for many genes previously implicated only by clinical or preclinical evidence, and it uncovers a plethora of novel disease genes with compelling biological rationale.

A multimodal deep learning model to infer cell-type-specific functional gene networks

BACKGROUND

Functional gene networks (FGNs) capture functional relationships among genes that vary across tissues and cell types. Construction of cell-type-specific FGNs enables the understanding of cell-type-specific functional gene relationships and insights into genetic mechanisms of human diseases in disease-relevant cell types. However, most existing FGNs were developed without consideration of specific cell types within tissues.

RESULTS

In this study, we created a multimodal deep learning model (MDLCN) to predict cell-type-specific FGNs in the human brain by integrating single-nuclei gene expression data with global protein interaction networks. We systematically evaluated the prediction performance of the MDLCN and showed its superior performance compared to two baseline models (boosting tree and convolutional neural network). Based on the predicted cell-type-specific FGNs, we observed that cell-type marker genes had a higher level of hubness than non-marker genes in their corresponding cell type. Furthermore, we showed that risk genes underlying autism and Alzheimer's disease were more strongly connected in disease-relevant cell types, supporting the cellular context of predicted cell-type-specific FGNs.

CONCLUSIONS

Our study proposes a powerful deep learning approach (MDLCN) to predict FGNs underlying a diverse set of cell types in human brain. The MDLCN model enhances prediction accuracy of cell-type-specific FGNs compared to single modality convolutional neural network (CNN) and boosting tree models, as shown by higher areas under both receiver operating characteristic (ROC) and precision-recall curves for different levels of independent test datasets. The predicted FGNs also show evidence for the cellular context and distinct topological features (i.e. higher hubness and topological score) of cell-type marker genes. Moreover, we observed stronger modularity among disease-associated risk genes in FGNs of disease-relevant cell types. For example, the strength of connectivity among autism risk genes was stronger in neurons, but risk genes underlying Alzheimer's disease were more connected in microglia.

The first genome-wide association study of internet addiction; Revealed substantial shared risk factors with neurodevelopmental psychiatric disorders

BACKGROUND

Internet addiction disorder (IAD) is listed as a disorder requiring further studies in the diagnostic and statistical manual of mental disorders (DSM-V). Psychological studies showed significant co-morbidity of IAD with depression, alcohol abuse, and anxiety disorder. Etiology and genetic bases of IAD are unclear.

AIMS

Present study aimed to investigate the genetic, psychological, and cognitive bases of a tendency to internet addiction.

METHODS AND PROCEDURES

DNA was extracted from blood samples of IADs (N = 16,520) and 18,000 matched non-psychiatric subjects. Genotyping for the subjects was performed using SNP Array. Psychological, neuropsychological, and neurological characteristics were conducted.

OUTCOMES AND RESULTS

Seventy-two SNPs in 24 genes have been detected significantly associated with IAD. Most of these SNPs were risk factors for psychiatric disorders. Most similarity detected with autism spectrum disorder, bipolar disorder and schizophrenia. Higher anxiety, stress, and neuroticism and deficits in working memory, attention, planning, and processing speed were detected in IADs.

CONCLUSIONS

This study is the first genome-wide association study of IAD that showed strong shared genetic bases with neurodevelopmental disabilities and psychiatric disorders.

IMPLICATIONS

Genetic risk factors in IADs may cause several cognitive and neurodevelopmental brain function abnormalities, which lead to excessive Internet usage. It may suggest that IAD could be a marker for vulnerability to developmental psychiatric disorders.

Genome-wide meta-analyses reveal novel loci for verbal short-term memory and learning

Understanding the genomic basis of memory processes may help in combating neurodegenerative disorders. Hence, we examined the associations of common genetic variants with verbal short-term memory and verbal learning in adults without dementia or stroke (N = 53,637). We identified novel loci in the intronic region of CDH18, and at 13q21 and 3p21.1, as well as an expected signal in the APOE/APOC1/TOMM40 region. These results replicated in an independent sample. Functional and bioinformatic analyses supported many of these loci and further implicated POC1. We showed that polygenic score for verbal learning associated with brain activation in right parieto-occipital region during working memory task. Finally, we showed genetic correlations of these memory traits with several neurocognitive and health outcomes. Our findings suggest a role of several genomic loci in verbal memory processes.

A novel candidate disease gene prioritization method using deep graph convolutional networks and semi-supervised learning

BACKGROUND

Selecting and prioritizing candidate disease genes is necessary before conducting laboratory studies as identifying disease genes from a large number of candidate genes using laboratory methods, is a very costly and time-consuming task. There are many machine learning-based gene prioritization methods. These methods differ in various aspects including the feature vectors of genes, the used datasets with different structures, and the learning model. Creating a suitable feature vector for genes and an appropriate learning model on a variety of data with different and non-Euclidean structures, including graphs, as well as the lack of negative data are very important challenges of these methods. The use of graph neural networks has recently emerged in machine learning and other related fields, and they have demonstrated superior performance for a broad range of problems.

METHODS

In this study, a new semi-supervised learning method based on graph convolutional networks is presented using the novel constructing feature vector for each gene. In the proposed method, first, we construct three feature vectors for each gene using terms from the Gene Ontology (GO) database. Then, we train a graph convolution network on these vectors using protein-protein interaction (PPI) network data to identify disease candidate genes. Our model discovers hidden layer representations encoding in both local graph structure as well as features of nodes. This method is characterized by the simultaneous consideration of topological information of the biological network (e.g., PPI) and other sources of evidence. Finally, a validation has been done to demonstrate the efficiency of our method.

RESULTS

Several experiments are performed on 16 diseases to evaluate the proposed method's performance. The experiments demonstrate that our proposed method achieves the best results, in terms of precision, the area under the ROC curve (AUCs), and F1-score values, when compared with eight state-of-the-art network and machine learning-based disease gene prioritization methods.

CONCLUSION

This study shows that the proposed semi-supervised learning method appropriately classifies and ranks candidate disease genes using a graph convolutional network and an innovative method to create three feature vectors for genes based on the molecular function, cellular component, and biological process terms from GO data.

TGFβ superfamily signaling regulates the state of human stem cell pluripotency and capacity to create well-structured telencephalic organoids

Telencephalic organoids generated from human pluripotent stem cells (hPSCs) are a promising system for studying the distinct features of the developing human brain and the underlying causes of many neurological disorders. While organoid technology is steadily advancing, many challenges remain, including potential batch-to-batch and cell-line-to-cell-line variability, and structural inconsistency. Here, we demonstrate that a major contributor to cortical organoid quality is the way hPSCs are maintained prior to differentiation. Optimal results were achieved using particular fibroblast-feeder-supported hPSCs rather than feeder-independent cells, differences that were reflected in their transcriptomic states at the outset. Feeder-supported hPSCs displayed activation of diverse transforming growth factor β (TGFβ) superfamily signaling pathways and increased expression of genes connected to naive pluripotency. We further identified combinations of TGFβ-related growth factors that are necessary and together sufficient to impart broad telencephalic organoid competency to feeder-free hPSCs and enhance the formation of well-structured brain tissues suitable for disease modeling.

Chr21 protein-protein interactions: enrichment in proteins involved in intellectual disability, autism, and late-onset Alzheimer's disease

Down syndrome (DS) is caused by human chromosome 21 (HSA21) trisomy. It is characterized by a poorly understood intellectual disability (ID). We studied two mouse models of DS, one with an extra copy of the <i>Dyrk1A</i> gene (189N3) and the other with an extra copy of the mouse Chr16 syntenic region (Dp(16)1Yey). RNA-seq analysis of the transcripts deregulated in the embryonic hippocampus revealed an enrichment in genes associated with chromatin for the 189N3 model, and synapses for the Dp(16)1Yey model. A large-scale yeast two-hybrid screen (82 different screens, including 72 HSA21 baits and 10 rebounds) of a human brain library containing at least 10⁷ independent fragments identified 1,949 novel protein-protein interactions. The direct interactors of HSA21 baits and rebounds were significantly enriched in ID-related genes (<i>P</i>-value &lt; 2.29 × 10^-8). Proximity ligation assays showed that some of the proteins encoded by HSA21 were located at the dendritic spine postsynaptic density, in a protein network at the dendritic spine postsynapse. We located HSA21 DYRK1A and DSCAM, mutations of which increase the risk of autism spectrum disorder (ASD) 20-fold, in this postsynaptic network. We found that an intracellular domain of DSCAM bound either DLGs, which are multimeric scaffolds comprising receptors, ion channels and associated signaling proteins, or DYRK1A. The DYRK1A-DSCAM interaction domain is conserved in <i>Drosophila</i> and humans. The postsynaptic network was found to be enriched in proteins associated with ARC-related synaptic plasticity, ASD, and late-onset Alzheimer's disease. These results highlight links between DS and brain diseases with a complex genetic basis.

Transcription factor network analysis identifies REST/NRSF as an intrinsic regulator of CNS regeneration in mice

The inability of neurons to regenerate long axons within the CNS is a major impediment to improving outcome after spinal cord injury, stroke, and other CNS insults. Recent advances have uncovered an intrinsic program that involves coordinate regulation by multiple transcription factors that can be manipulated to enhance growth in the peripheral nervous system. Here, we use a systems genomics approach to characterize regulatory relationships of regeneration-associated transcription factors, identifying RE1-Silencing Transcription Factor (REST; Neuron-Restrictive Silencer Factor, NRSF) as a predicted upstream suppressor of a pro-regenerative gene program associated with axon regeneration in the CNS. We validate our predictions using multiple paradigms, showing that mature mice bearing cell type-specific deletions of REST or expressing dominant-negative mutant REST show improved regeneration of the corticospinal tract and optic nerve after spinal cord injury and optic nerve crush, which is accompanied by upregulation of regeneration-associated genes in cortical motor neurons and retinal ganglion cells, respectively. These analyses identify a role for REST as an upstream suppressor of the intrinsic regenerative program in the CNS and demonstrate the utility of a systems biology approach involving integrative genomics and bio-informatics to prioritize hypotheses relevant to CNS repair.

The 22q11.2 region regulates presynaptic gene-products linked to schizophrenia

It is unclear how the 22q11.2 deletion predisposes to psychiatric disease. To study this, we generated induced pluripotent stem cells from deletion carriers and controls and utilized CRISPR/Cas9 to introduce the heterozygous deletion into a control cell line. Here, we show that upon differentiation into neural progenitor cells, the deletion acted in trans to alter the abundance of transcripts associated with risk for neurodevelopmental disorders including autism. In excitatory neurons, altered transcripts encoded presynaptic factors and were associated with genetic risk for schizophrenia, including common and rare variants. To understand how the deletion contributed to these changes, we defined the minimal protein-protein interaction network that best explains gene expression alterations. We found that many genes in 22q11.2 interact in presynaptic, proteasome, and JUN/FOS transcriptional pathways. Our findings suggest that the 22q11.2 deletion impacts genes that may converge with psychiatric risk loci to influence disease manifestation in each deletion carrier.

Network-Based Methods for Approaching Human Pathologies from a Phenotypic Point of View

Network and systemic approaches to studying human pathologies are helping us to gain insight into the molecular mechanisms of and potential therapeutic interventions for human diseases, especially for complex diseases where large numbers of genes are involved. The complex human pathological landscape is traditionally partitioned into discrete “diseases”; however, that partition is sometimes problematic, as diseases are highly heterogeneous and can differ greatly from one patient to another. Moreover, for many pathological states, the set of symptoms (phenotypes) manifested by the patient is not enough to diagnose a particular disease. On the contrary, phenotypes, by definition, are directly observable and can be closer to the molecular basis of the pathology. These clinical phenotypes are also important for personalised medicine, as they can help stratify patients and design personalised interventions. For these reasons, network and systemic approaches to pathologies are gradually incorporating phenotypic information. This review covers the current landscape of phenotype-centred network approaches to study different aspects of human diseases.

Different responses to risperidone treatment in Schizophrenia: a multicenter genome-wide association and whole exome sequencing joint study

Risperidone is routinely used in the clinical management of schizophrenia, but the treatment response is highly variable among different patients. The genetic underpinnings of the treatment response are not well understood. We performed a pharmacogenomic study of the treatment response to risperidone in patients with schizophrenia by using a SNP microarray -based genome-wide association study (GWAS) and whole exome sequencing (WES)-based GWAS. DNA samples were collected from 189 patients for the GWAS and from 222 patients for the WES after quality control in multiple centers of China. Antipsychotic response phenotypes of patients who received eight weeks of risperidone treatment were quantified with percentage change on the Positive and Negative Syndrome Scale (PANSS). The GWAS revealed a significant association between several SNPs and treatment response, such as three GRM7 SNPs (rs141134664, rs57521140, and rs73809055). Gene-based analysis in WES revealed 13 genes that were associated with antipsychotic response, such as GPR12 and MAP2K3. We did not identify shared loci or genes between GWAS and WES, but association signals tended to cluster into the GPCR gene family and GPCR signaling pathway, which may play an important role in the treatment response etiology. This study may provide a research paradigm for pharmacogenomic research, and these data provide a promising illustration of our potential to identify genetic variants underlying antipsychotic responses and may ultimately facilitate precision medicine in schizophrenia.

Molecular profiles and urinary biomarkers of upper tract urothelial carcinomas associated with aristolochic acid exposure

Recurrent upper tract urothelial carcinomas (UTUCs) arise in the context of nephropathy linked to exposure to the herbal carcinogen aristolochic acid (AA). Here we delineated the molecular programs underlying UTUC tumorigenesis in patients from endemic aristolochic acid nephropathy (AAN) regions in Southern Europe. We applied an integrative multiomics analysis of UTUCs, corresponding unaffected tissues and of patient urines. Quantitative microRNA (miRNA) and messenger ribonucleic acid (mRNA) expression profiling, immunohistochemical analysis by tissue microarrays and exome and transcriptome sequencing were performed in UTUC and nontumor tissues. Urinary miRNAs of cases undergoing surgery were profiled before and after tumor resection. Ribonucleic acid (RNA) and protein levels were analyzed using appropriate statistical tests and trend assessment. Dedicated bioinformatic tools were used for analysis of pathways, mutational signatures and result visualization. The results delineate UTUC-specific miRNA:mRNA networks comprising 89 miRNAs associated with 1,862 target mRNAs, involving deregulation of cell cycle, deoxyribonucleic acid (DNA) damage response, DNA repair, bladder cancer, oncogenes, tumor suppressors, chromatin structure regulators and developmental signaling pathways. Key UTUC-specific transcripts were confirmed at the protein level. Exome and transcriptome sequencing of UTUCs revealed AA-specific mutational signature SBS22, with 68% to 76% AA-specific, deleterious mutations propagated at the transcript level, a possible basis for neoantigen formation and immunotherapy targeting. We next identified a signature of UTUC-specific miRNAs consistently more abundant in the patients' urine prior to tumor resection, thereby defining biomarkers of tumor presence. The complex gene regulation programs of AAN-associated UTUC tumors involve regulatory miRNAs prospectively applicable to noninvasive urine-based screening of AAN patients for cancer presence and recurrence.

DeepSVP: integration of genotype and phenotype for structural variant prioritization using deep learning

MOTIVATION

Structural genomic variants account for much of human variability and are involved in several diseases. Structural variants are complex and may affect coding regions of multiple genes, or affect the functions of genomic regions in different ways from single nucleotide variants. Interpreting the phenotypic consequences of structural variants relies on information about gene functions, haploinsufficiency or triplosensitivity and other genomic features. Phenotype-based methods to identifying variants that are involved in genetic diseases combine molecular features with prior knowledge about the phenotypic consequences of altering gene functions. While phenotype-based methods have been applied successfully to single nucleotide variants as well as short insertions and deletions, the complexity of structural variants makes it more challenging to link them to phenotypes. Furthermore, structural variants can affect a large number of coding regions, and phenotype information may not be available for all of them.

RESULTS

We developed DeepSVP, a computational method to prioritize structural variants involved in genetic diseases by combining genomic and gene functions information. We incorporate phenotypes linked to genes, functions of gene products, gene expression in individual cell types and anatomical sites of expression, and systematically relate them to their phenotypic consequences through ontologies and machine learning. DeepSVP significantly improves the success rate of finding causative variants in several benchmarks and can identify novel pathogenic structural variants in consanguineous families.

AVAILABILITY AND IMPLEMENTATION

https://github.com/bio-ontology-research-group/DeepSVP.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A candidate biological network formed by genes from genomic and hypothesis-free scans of suicide

Information about genes and the biology of suicidal behavior (SB) is noisy due to heterogenous outcomes (suicide attempts or deaths), as well as many different genes and overlapping biological processes implicated. One approach to test the unbiased biological coherence of disease genes, is to use genes from hypothesis-free genetic scans and to investigate if they aggregate close to each other in cellular gene and protein interaction networks ("interactomes"). Therefore, we used network methods to study the biological coherence among genes (n = 229) from genome-wide association studies (GWAS) and whole exome sequencing (WES) of suicide outcome. Results showed that the suicide GWAS+WES genes has significant aggregation in three major interactome database assemblies, a hallmark of biological similarity and increased likelihood of being involved in the same outcome (suicide). This pinpointed e.g. genes on chromosome 19, which are also associated with lipid metabolism, schizophrenia and bipolar disorder. We identified a subset of GWAS+WES "core" genes (n = 54) which are the most proximal to each other in the context of three interactome assemblies, and present a candidate network module of suicide which is specific for nervous system tissues. The n = 54 most proximal "core" genes showed overrepresentation of synaptic and nervous system development genes, as well as network paths to other SB genes having increased evidence diversity. Overall, results suggested the existence of a coherent biology in suicide outcome and provide unbiased biological support concerning links to other SB genes, as well as e.g. bipolar disorder, excitatory/inhibitory function and ketamine treatment in SB.

Prioritization of disease genes from GWAS using ensemble-based positive-unlabeled learning

A primary challenge in understanding disease biology from genome-wide association studies (GWAS) arises from the inability to directly implicate causal genes from association data. Integration of multiple-omics data sources potentially provides important functional links between associated variants and candidate genes. Machine-learning is well-positioned to take advantage of a variety of such data and provide a solution for the prioritization of disease genes. Yet, classical positive-negative classifiers impose strong limitations on the gene prioritization procedure, such as a lack of reliable non-causal genes for training. Here, we developed a novel gene prioritization tool-Gene Prioritizer (GPrior). It is an ensemble of five positive-unlabeled bagging classifiers (Logistic Regression, Support Vector Machine, Random Forest, Decision Tree, Adaptive Boosting), that treats all genes of unknown relevance as an unlabeled set. GPrior selects an optimal composition of algorithms to tune the model for each specific phenotype. Altogether, GPrior fills an important niche of methods for GWAS data post-processing, significantly improving the ability to pinpoint disease genes compared to existing solutions.

Unveiling the Pathogenesis of Psychiatric Disorders Using Network Models

Psychiatric disorders are complex brain disorders with a high degree of genetic heterogeneity, affecting millions of people worldwide. Despite advances in psychiatric genetics, the underlying pathogenic mechanisms of psychiatric disorders are still largely elusive, which impedes the development of novel rational therapies. There has been accumulating evidence suggesting that the genetics of complex disorders can be viewed through an omnigenic lens, which involves contextualizing genes in highly interconnected networks. Thus, applying network-based multi-omics integration methods could cast new light on the pathophysiology of psychiatric disorders. In this review, we first provide an overview of the recent advances in psychiatric genetics and highlight gaps in translating molecular associations into mechanistic insights. We then present an overview of network methodologies and review previous applications of network methods in the study of psychiatric disorders. Lastly, we describe the potential of such methodologies within a multi-tissue, multi-omics approach, and summarize the future directions in adopting diverse network approaches.

Neuroimaging PheWAS (Phenome-Wide Association Study): A Free Cloud-Computing Platform for Big-Data, Brain-Wide Imaging Association Studies

Large-scale, case-control genome-wide association studies (GWASs) have revealed genetic variations associated with diverse neurological and psychiatric disorders. Recent advances in neuroimaging and genomic databases of large healthy and diseased cohorts have empowered studies to characterize effects of the discovered genetic factors on brain structure and function, implicating neural pathways and genetic mechanisms in the underlying biology. However, the unprecedented scale and complexity of the imaging and genomic data requires new advanced biomedical data science tools to manage, process and analyze the data. In this work, we introduce Neuroimaging PheWAS (phenome-wide association study): a web-based system for searching over a wide variety of brain-wide imaging phenotypes to discover true system-level gene-brain relationships using a unified genotype-to-phenotype strategy. This design features a user-friendly graphical user interface (GUI) for anonymous data uploading, study definition and management, and interactive result visualizations as well as a cloud-based computational infrastructure and multiple state-of-art methods for statistical association analysis and multiple comparison correction. We demonstrated the potential of Neuroimaging PheWAS with a case study analyzing the influences of the apolipoprotein E (APOE) gene on various brain morphological properties across the brain in the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort. Benchmark tests were performed to evaluate the system's performance using data from UK Biobank. The Neuroimaging PheWAS system is freely available. It simplifies the execution of PheWAS on neuroimaging data and provides an opportunity for imaging genetics studies to elucidate routes at play for specific genetic variants on diseases in the context of detailed imaging phenotypic data.

Genoppi is an open-source software for robust and standardized integration of proteomic and genetic data

Combining genetic and cell-type-specific proteomic datasets can generate biological insights and therapeutic hypotheses, but a technical and statistical framework for such analyses is lacking. Here, we present an open-source computational tool called Genoppi (lagelab.org/genoppi) that enables robust, standardized, and intuitive integration of quantitative proteomic results with genetic data. We use Genoppi to analyze 16 cell-type-specific protein interaction datasets of four proteins (BCL2, TDP-43, MDM2, PTEN) involved in cancer and neurological disease. Through systematic quality control of the data and integration with published protein interactions, we show a general pattern of both cell-type-independent and cell-type-specific interactions across three cancer cell types and one human iPSC-derived neuronal cell type. Furthermore, through the integration of proteomic and genetic datasets in Genoppi, our results suggest that the neuron-specific interactions of these proteins are mediating their genetic involvement in neurodegenerative diseases. Importantly, our analyses suggest that human iPSC-derived neurons are a relevant model system for studying the involvement of BCL2 and TDP-43 in amyotrophic lateral sclerosis.

Transcriptomic networks implicate neuronal energetic abnormalities in three mouse models harboring autism and schizophrenia-associated mutations

Genetic risk for psychiatric illness is complex, so identification of shared molecular pathways where distinct forms of genetic risk might coincide is of substantial interest. A growing body of genetic and genomic studies suggest that such shared molecular pathways exist across disorders with different clinical presentations, such as schizophrenia and autism spectrum disorder (ASD). But how this relates to specific genetic risk factors is unknown. Further, whether some of the molecular changes identified in brain relate to potentially confounding antemortem or postmortem factors are difficult to prove. We analyzed the transcriptome from the cortex and hippocampus of three mouse lines modeling human copy number variants (CNVs) associated with schizophrenia and ASD: Df(h15q13)/+, Df(h22q11)/+, and Df(h1q21)/+ which carry the 15q13.3 deletion, 22q11.2 deletion, and 1q21.1 deletion, respectively. Although we found very little overlap of differential expression at the level of individual genes, gene network analysis identified two cortical and two hippocampal modules of co-expressed genes that were dysregulated across all three mouse models. One cortical module was associated with neuronal energetics and firing rate, and overlapped with changes identified in postmortem human brain from SCZ and ASD patients. These data highlight aspects of convergent gene expression in mouse models harboring major risk alleles, and strengthen the connection between changes in neuronal energetics and neuropsychiatric disorders in humans.

Single-Cell RNA Sequencing in Parkinson's Disease

Single-cell and single-nucleus RNA sequencing (sc/snRNA-seq) technologies have enhanced the understanding of the molecular pathogenesis of neurodegenerative disorders, including Parkinson's disease (PD). Nonetheless, their application in PD has been limited due mainly to the technical challenges resulting from the scarcity of postmortem brain tissue and low quality associated with RNA degradation. Despite such challenges, recent advances in animals and human in vitro models that recapitulate features of PD along with sequencing assays have fueled studies aiming to obtain an unbiased and global view of cellular composition and phenotype of PD at the single-cell resolution. Here, we reviewed recent sc/snRNA-seq efforts that have successfully characterized diverse cell-type populations and identified cell type-specific disease associations in PD. We also examined how these studies have employed computational and analytical tools to analyze and interpret the rich information derived from sc/snRNA-seq. Finally, we highlighted important limitations and emerging technologies for addressing key technical challenges currently limiting the integration of new findings into clinical practice.

Gene co-expression network analysis in human spinal cord highlights mechanisms underlying amyotrophic lateral sclerosis susceptibility

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease defined by motor neuron (MN) loss. Multiple genetic risk factors have been identified, implicating RNA and protein metabolism and intracellular transport, among other biological mechanisms. To achieve a systems-level understanding of the mechanisms governing ALS pathophysiology, we built gene co-expression networks using RNA-sequencing data from control human spinal cord samples, identifying 13 gene co-expression modules, each of which represents a distinct biological process or cell type. Analysis of four RNA-seq datasets from a range of ALS disease-associated contexts reveal dysregulation in numerous modules related to ribosomal function, wound response, and leukocyte activation, implicating astrocytes, oligodendrocytes, endothelia, and microglia in ALS pathophysiology. To identify potentially causal processes, we partitioned heritability across the genome, finding that ALS common genetic risk is enriched within two specific modules, SC.M4, representing genes related to RNA processing and gene regulation, and SC.M2, representing genes related to intracellular transport and autophagy and enriched in oligodendrocyte markers. Top hub genes of this latter module include ALS-implicated risk genes such as KPNA3, TMED2, and NCOA4, the latter of which regulates ferritin autophagy, implicating this process in ALS pathophysiology. These unbiased, genome-wide analyses confirm the utility of a systems approach to understanding the causes and drivers of ALS.

Core transcriptional networks in Williams syndrome: IGF1-PI3K-AKT-mTOR, MAPK and actin signaling at the synapse echo autism

Gene networks for disorders of social behavior provide the mechanisms critical for identifying therapeutic targets and biomarkers. Large behavioral phenotypic effects of small human deletions make the positive sociality of Williams syndrome (WS) ideal for determining transcriptional networks for social dysfunction currently based on DNA variations for disorders such as autistic spectrum disorder (ASD) and schizophrenia (SCHZ). Consensus on WS networks has been elusive due to the need for larger cohort size, sensitive genome-wide detection and analytic tools. We report a core set of WS network perturbations in a cohort of 58 individuals (34 with typical, 6 atypical deletions and 18 controls). Genome-wide exon-level expression arrays robustly detected changes in differentially expressed gene (DEG) transcripts from WS deleted genes that ranked in the top 11 of 12 122 transcripts, validated by quantitative reverse transcription PCR, RNASeq and western blots. WS DEG's were strictly dosed in the full but not the atypical deletions that revealed a breakpoint position effect on non-deleted CLIP2, a caveat for current phenotypic mapping based on copy number variants. Network analyses tested the top WS DEG's role in the dendritic spine, employing GeneMANIA to harmonize WS DEGs with comparable query gene-sets. The results indicate perturbed actin cytoskeletal signaling analogous to the excitatory dendritic spines. Independent protein-protein interaction analyses of top WS DEGs generated a 100-node graph annotated topologically revealing three interacting pathways, MAPK, IGF1-PI3K-AKT-mTOR/insulin and actin signaling at the synapse. The results indicate striking similarity of WS transcriptional networks to genome-wide association study-based ASD and SCHZ risk suggesting common network dysfunction for these disorders of divergent sociality.

Gene- and pathway-level analyses of iCOGS variants highlight novel signaling pathways underlying familial breast cancer susceptibility

Single‐nucleotide polymorphisms (SNPs) in over 180 loci have been associated with breast cancer (BC) through genome‐wide association studies involving mostly unselected population‐based case‐control series. Some of them modify BC risk of women carrying a BRCA1 or BRCA2 (BRCA1/2) mutation and may also explain BC risk variability in BC‐prone families with no BRCA1/2 mutation. Here, we assessed the contribution of SNPs of the iCOGS array in GENESIS consisting of BC cases with no BRCA1/2 mutation and a sister with BC, and population controls. Genotyping data were available for 1281 index cases, 731 sisters with BC, 457 unaffected sisters and 1272 controls. In addition to the standard SNP‐level analysis using index cases and controls, we performed pedigree‐based association tests to capture transmission information in the sibships. We also performed gene‐ and pathway‐level analyses to maximize the power to detect associations with lower‐frequency SNPs or those with modest effect sizes. While SNP‐level analyses identified 18 loci, gene‐level analyses identified 112 genes. Furthermore, 31 Kyoto Encyclopedia of Genes and Genomes and 7 Atlas of Cancer Signaling Network pathways were highlighted (false discovery rate of 5%). Using results from the “index case‐control” analysis, we built pathway‐derived polygenic risk scores (PRS) and assessed their performance in the population‐based CECILE study and in a data set composed of GENESIS‐affected sisters and CECILE controls. Although these PRS had poor predictive value in the general population, they performed better than a PRS built using our SNP‐level findings, and we found that the joint effect of family history and PRS needs to be considered in risk prediction models.

What's new?

Genetic studies have identified more than 180 single‐nucleotide polymorphisms (SNPs) associated with breast cancer susceptibility, but these studies are reaching their limits. Here, the authors evaluated SNPs in the iCOGS genotyping array using a multilevel approach, including single variant, gene, and pathway analyses. They measured the contribution of the SNPs to breast cancer in patients who have a sister with breast cancer but do not carry a BRCA1/2 mutation. They showed that a pathway‐derived polygenic risk score performed poorly in the general population, and that the best predictive model must include family history.

A Combined Analysis of Genetically Correlated Traits Identifies Genes and Brain Regions for Insomnia

AIMS

Previous studies have inferred that there is a strong genetic component in insomnia. However, the etiology of insomnia is still unclear. This study systematically analyzed multiple genome-wide association study (GWAS) data sets with core human pathways and functional networks to detect potential gene pathways and networks associated with insomnia.

METHODS

We used a novel method, multitrait analysis of genome-wide association studies (MTAG), to combine 3 large GWASs of insomnia symptoms/complaints and sleep duration. The i-Gsea4GwasV2 and Reactome FI programs were used to analyze data from the result of MTAG analysis and the nominally significant pathways, respectively.

RESULTS

Through analyzing data sets using the MTAG program, our sample size increased from 113,006 subjects to 163,188 subjects. A total of 17 of 1,816 Reactome pathways were identified and showed to be associated with insomnia. We further revealed 11 interconnected functional and topologically interacting clusters (Clusters 0 to 10) that were associated with insomnia. Based on the brain transcriptome data, it was found that the genes in Cluster 4 were enriched for the transcriptional coexpression profile in the prenatal dorsolateral prefrontal cortex (P = 7 × 10-5), inferolateral temporal cortex (P = 0.02), medial prefrontal cortex (P < 1 × 10-5), and amygdala (P < 1 × 10-5), and detected RPA2, ORC6, PIAS3, and PRIM2 as core nodes in these 4 brain regions.

CONCLUSIONS

The findings provided new genes, pathways, and brain regions to understand the pathology of insomnia.

TWAS pathway method greatly enhances the number of leads for uncovering the molecular underpinnings of psychiatric disorders

Genetic signal detection in genome-wide association studies (GWAS) is enhanced by pooling small signals from multiple Single Nucleotide Polymorphism (SNP), for example, across genes and pathways. Because genes are believed to influence traits via gene expression, it is of interest to combine information from expression Quantitative Trait Loci (eQTLs) in a gene or genes in the same pathway. Such methods, widely referred to as transcriptomic wide association studies (TWAS), already exist for gene analysis. Due to the possibility of eliminating most of the confounding effects of linkage disequilibrium (LD) from TWAS gene statistics, pathway TWAS methods would be very useful in uncovering the true molecular basis of psychiatric disorders. However, such methods are not yet available for arbitrarily large pathways/gene sets. This is possibly due to the quadratic (as a function of the number of SNPs) computational burden for computing LD across large chromosomal regions. To overcome this obstacle, we propose JEPEGMIX2-P, a novel TWAS pathway method that (a) has a linear computational burden, (b) uses a large and diverse reference panel (33 K subjects), (c) is competitive (adjusts for background enrichment in gene TWAS statistics), and (d) is applicable as-is to ethnically mixed-cohorts. To underline its potential for increasing the power to uncover genetic signals over the commonly used nontranscriptomics methods, for example, MAGMA, we applied JEPEGMIX2-P to summary statistics of most large meta-analyses from Psychiatric Genetics Consortium (PGC). While our work is just the very first step toward clinical translation of psychiatric disorders, PGC anorexia results suggest a possible avenue for treatment.

Markov chain Monte Carlo for active module identification problem

Background

Integrative network methods are commonly used for interpretation of high-throughput experimental biological data: transcriptomics, proteomics, metabolomics and others. One of the common approaches is finding a connected subnetwork of a global interaction network that best encompasses significant individual changes in the data and represents a so-called active module. Usually methods implementing this approach find a single subnetwork and thus solve a hard classification problem for vertices. This subnetwork inherently contains erroneous vertices, while no instrument is provided to estimate the confidence level of any particular vertex inclusion. To address this issue, in the current study we consider the active module problem as a soft classification problem.

Results

We propose a method to estimate probabilities of each vertex to belong to the active module based on Markov chain Monte Carlo (MCMC) subnetwork sampling. As an example of the performance of our method on real data, we run it on two gene expression datasets. For the first many-replicate expression dataset we show that the proposed approach is consistent with an existing resampling-based method. On the second dataset the jackknife resampling method is inapplicable due to the small number of biological replicates, but the MCMC method can be run and shows high classification performance.

Conclusions

The proposed method allows to estimate the probability that an individual vertex belongs to the active module as well as the false discovery rate (FDR) for a given set of vertices. Given the estimated probabilities, it becomes possible to provide a connected subgraph in a consistent manner for any given FDR level: no vertex can disappear when the FDR level is relaxed. We show, on both simulated and real datasets, that the proposed method has good computational performance and high classification accuracy.

Tau Pathology Drives Dementia Risk-Associated Gene Networks toward Chronic Inflammatory States and Immunosuppression

To understand how neural-immune-associated genes and pathways contribute to neurodegenerative disease pathophysiology, we performed a systematic functional genomic analysis in purified microglia and bulk tissue from mouse and human AD, FTD, and PSP. We uncover a complex temporal trajectory of microglial-immune pathways involving the type 1 interferon response associated with tau pathology in the early stages, followed by later signatures of partial immune suppression and, subsequently, the type 2 interferon response. We find that genetic risk for dementias shows disease-specific patterns of pathway enrichment. We identify drivers of two gene co-expression modules conserved from mouse to human, representing competing arms of microglial-immune activation (NAct) and suppression (NSupp) in neurodegeneration. We validate our findings by using chemogenetics, experimental perturbation data, and single-cell sequencing in post-mortem brains. Our results refine the understanding of stage- and disease-specific microglial responses, implicate microglial viral defense pathways in dementia pathophysiology, and highlight therapeutic windows.

Rexach et al. use transcriptional network analysis to define dynamic microglial transitions across neurodegeneration, discovering that three dementias with tau pathology involve dysregulated microglial viral and antiviral pathways. Bio-informatics coupled with experimental validation identifies regulatory drivers, implicating double-stranded RNA and interferon-response genes as drivers of early immune suppression in disease.

A systems genomics approach to uncover the molecular properties of cancer genes

Genes involved in cancer are under constant evolutionary pressure, potentially resulting in diverse molecular properties. In this study, we explore 23 omic features from publicly available databases to define the molecular profile of different classes of cancer genes. Cancer genes were grouped according to mutational landscape (germline and somatically mutated genes), role in cancer initiation (cancer driver genes) or cancer survival (survival genes), as well as being implicated by genome-wide association studies (GWAS genes). For each gene, we also computed feature scores based on all omic features, effectively summarizing how closely a gene resembles cancer genes of the respective class. In general, cancer genes are longer, have a lower GC content, have more isoforms with shorter exons, are expressed in more tissues and have more transcription factor binding sites than non-cancer genes. We found that germline genes more closely resemble single tissue GWAS genes while somatic genes are more similar to pleiotropic cancer GWAS genes. As a proof-of-principle, we utilized aggregated feature scores to prioritize genes in breast cancer GWAS loci and found that top ranking genes were enriched in cancer related pathways. In conclusion, we have identified multiple omic features associated with different classes of cancer genes, which can assist prioritization of genes in cancer gene discovery.

Integrative genomics approach identifies conserved transcriptomic networks in Alzheimer's disease

Alzheimer's disease (AD) is a devastating neurological disorder characterized by changes in cell-type proportions and consequently marked alterations of the transcriptome. Here we use a data-driven systems biology meta-analytical approach across three human AD cohorts, encompassing six cortical brain regions, and integrate with multi-scale datasets comprising of DNA methylation, histone acetylation, transcriptome- and genome-wide association studies and quantitative trait loci to further characterize the genetic architecture of AD. We perform co-expression network analysis across more than 1200 human brain samples, identifying robust AD-associated dysregulation of the transcriptome, unaltered in normal human aging. We assess the cell-type specificity of AD gene co-expression changes and estimate cell-type proportion changes in human AD by integrating co-expression modules with single-cell transcriptome data generated from 27 321 nuclei from human postmortem prefrontal cortical tissue. We also show that genetic variants of AD are enriched in a microglial AD-associated module and identify key transcription factors regulating co-expressed modules. Additionally, we validate our results in multiple published human AD gene expression datasets, which can be easily accessed using our online resource (https://swaruplab.bio.uci.edu/consensusAD).

A Machine Learning Approach to Predicting Autism Risk Genes: Validation of Known Genes and Discovery of New Candidates

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition with a strong genetic basis. The role of de novo mutations in ASD has been well established, but the set of genes implicated to date is still far from complete. The current study employs a machine learning-based approach to predict ASD risk genes using features from spatiotemporal gene expression patterns in human brain, gene-level constraint metrics, and other gene variation features. The genes identified through our prediction model were enriched for independent sets of ASD risk genes, and tended to be down-expressed in ASD brains, especially in frontal and parietal cortex. The highest-ranked genes not only included those with strong prior evidence for involvement in ASD (for example, NBEA, HERC1, and TCF20), but also indicated potentially novel candidates, such as, MYCBP2 and CAND1, which are involved in protein ubiquitination. We also showed that our method outperformed state-of-the-art scoring systems for ranking curated ASD candidate genes. Gene ontology enrichment analysis of our predicted risk genes revealed biological processes clearly relevant to ASD, including neuronal signaling, neurogenesis, and chromatin remodeling, but also highlighted other potential mechanisms that might underlie ASD, such as regulation of RNA alternative splicing and ubiquitination pathway related to protein degradation. Our study demonstrates that human brain spatiotemporal gene expression patterns and gene-level constraint metrics can help predict ASD risk genes. Our gene ranking system provides a useful resource for prioritizing ASD candidate genes.

From Rare Copy Number Variants to Biological Processes in ADHD

OBJECTIVE

Attention deficit hyperactivity disorder (ADHD) is a highly heritable psychiatric disorder. The objective of this study was to define ADHD-associated candidate genes and their associated molecular modules and biological themes, based on the analysis of rare genetic variants.

METHODS

The authors combined data from 11 published copy number variation studies in 6,176 individuals with ADHD and 25,026 control subjects and prioritized genes by applying an integrative strategy based on criteria including recurrence in individuals with ADHD, absence in control subjects, complete coverage in copy number gains, and presence in the minimal region common to overlapping copy number variants (CNVs), as well as on protein-protein interactions and information from cross-species genotype-phenotype annotation.

RESULTS

The authors localized 2,241 eligible genes in the 1,532 reported CNVs, of which they classified 432 as high-priority ADHD candidate genes. The high-priority ADHD candidate genes were significantly coexpressed in the brain. A network of 66 genes was supported by ADHD-relevant phenotypes in the cross-species database. Four significantly interconnected protein modules were found among the high-priority ADHD genes. A total of 26 genes were observed across all applied bioinformatic methods. Lookup in the latest genome-wide association study for ADHD showed that among those 26 genes, POLR3C and RBFOX1 were also supported by common genetic variants.

CONCLUSIONS

Integration of a stringent filtering procedure in CNV studies with suitable bioinformatics approaches can identify ADHD candidate genes at increased levels of credibility. The authors' analytic pipeline provides additional insight into the molecular mechanisms underlying ADHD and allows prioritization of genes for functional validation in validated model organisms.

Differential regulation of JAK/STAT-signaling in patients with ulcerative colitis and Crohn’s disease

In 2018, the pan-Janus kinase (JAK) inhibitor tofacitinib was launched for the treatment of ulcerative colitis (UC). Although tofacitinib has proven efficacious in patients with active UC, it failed in patients with Crohn’s disease (CD). This finding strongly hints at a different contribution of JAK signaling in both entities. Here, we review the current knowledge on the interplay between the JAK/signal transducer and activator of transcription (STAT) pathway and inflammatory bowel diseases (IBD). In particular, we provide a detailed overview of the differences and similarities of JAK/STAT-signaling in UC and CD, highlight the impact of the JAK/STAT pathway in experimental colitis models and summarize the published evidence on JAK/STAT-signaling in immune cells of IBD as well as the genetic association between the JAK/STAT pathway and IBD. Finally, we describe novel treatment strategies targeting JAK/STAT inhibition in UC and CD and comment on the limitations and challenges of the new drug class.

Protein-Protein interactions uncover candidate 'core genes' within omnigenic disease networks

Genome wide association studies (GWAS) of human diseases have generally identified many loci associated with risk with relatively small effect sizes. The omnigenic model attempts to explain this observation by suggesting that diseases can be thought of as networks, where genes with direct involvement in disease-relevant biological pathways are named ‘core genes’, while peripheral genes influence disease risk via their interactions or regulatory effects on core genes. Here, we demonstrate a method for identifying candidate core genes solely from genes in or near disease-associated SNPs (GWAS hits) in conjunction with protein-protein interaction network data. Applied to 1,381 GWAS studies from 5 ancestries, we identify a total of 1,865 candidate core genes in 343 GWAS studies. Our analysis identifies several well-known disease-related genes that are not identified by GWAS, including BRCA1 in Breast Cancer, Amyloid Precursor Protein (APP) in Alzheimer’s Disease, INS in A1C measurement and Type 2 Diabetes, and PCSK9 in LDL cholesterol, amongst others. Notably candidate core genes are preferentially enriched for disease relevance over GWAS hits and are enriched for both Clinvar pathogenic variants and known drug targets—consistent with the predictions of the omnigenic model. We subsequently use parent term annotations provided by the GWAS catalog, to merge related GWAS studies and identify candidate core genes in over-arching disease processes such as cancer–where we identify 109 candidate core genes.

A recent theory suggests that only a small number of genes underpin the biology of a disease, these genes are called ‘core genes’, and for most diseases, these core genes remain unknown. The suggested methods for finding them requires complex and expensive experiments. We reasoned that if we merge currently available datasets in smart ways, we may be able to uncover these ‘core genes’. Our method finds “hub” proteins by merging lists of genes previously linked with disease to information on how proteins interact with each other. We found that many of these hub proteins have central roles in disease, such as insulin for both A1C measurement and Type 2 Diabetes, BRCA1 in Breast cancer, and Amyloid Precursor Protein in Alzheimer’s Disease. We think these ‘hub’ proteins are candidate ‘core genes’, and offer our method as a way to find ‘core genes’ by utilizing publicly available reference datasets.

Synaptic and brain-expressed gene sets relate to the shared genetic risk across five psychiatric disorders

BACKGROUND

Mounting evidence shows genetic overlap between multiple psychiatric disorders. However, the biological underpinnings of shared risk for psychiatric disorders are not yet fully uncovered. The identification of underlying biological mechanisms is crucial for the progress in the treatment of these disorders.

METHODS

We applied gene-set analysis including 7372 gene sets, and 53 tissue-type specific gene-expression profiles to identify sets of genes that are involved in the etiology of multiple psychiatric disorders. We included genome-wide meta-association data of the five psychiatric disorders schizophrenia, bipolar disorder, major depressive disorder, autism spectrum disorder, and attention-deficit/hyperactivity disorder. The total dataset contained 159 219 cases and 262 481 controls.

RESULTS

We identified 19 gene sets that were significantly associated with the five psychiatric disorders combined, of which we excluded five sets because their associations were likely driven by schizophrenia only. Conditional analyses showed independent effects of several gene sets that in particular relate to the synapse. In addition, we found independent effects of gene expression levels in the cerebellum and frontal cortex.

CONCLUSIONS

We obtained novel evidence for shared biological mechanisms that act across psychiatric disorders and we showed that several gene sets that have been related to individual disorders play a role in a broader range of psychiatric disorders.

Prediction of cancer driver genes through network-based moment propagation of mutation scores

MOTIVATION

Gaining a comprehensive understanding of the genetics underlying cancer development and progression is a central goal of biomedical research. Its accomplishment promises key mechanistic, diagnostic and therapeutic insights. One major step in this direction is the identification of genes that drive the emergence of tumors upon mutation. Recent advances in the field of computational biology have shown the potential of combining genetic summary statistics that represent the mutational burden in genes with biological networks, such as protein-protein interaction networks, to identify cancer driver genes. Those approaches superimpose the summary statistics on the nodes in the network, followed by an unsupervised propagation of the node scores through the network. However, this unsupervised setting does not leverage any knowledge on well-established cancer genes, a potentially valuable resource to improve the identification of novel cancer drivers.

RESULTS

We develop a novel node embedding that enables classification of cancer driver genes in a supervised setting. The embedding combines a representation of the mutation score distribution in a node's local neighborhood with network propagation. We leverage the knowledge of well-established cancer driver genes to define a positive class, resulting in a partially labeled dataset, and develop a cross-validation scheme to enable supervised prediction. The proposed node embedding followed by a supervised classification improves the predictive performance compared with baseline methods and yields a set of promising genes that constitute candidates for further biological validation.

AVAILABILITY AND IMPLEMENTATION

Code available at https://github.com/BorgwardtLab/MoProEmbeddings.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Multi-trait analysis for genome-wide association study of five psychiatric disorders

We conducted a cross-trait meta-analysis of genome-wide association study on schizophrenia (SCZ) (n = 65,967), bipolar disorder (BD) (n = 41,653), autism spectrum disorder (ASD) (n = 46,350), attention deficit hyperactivity disorder (ADHD) (n = 55,374), and depression (DEP) (n = 688,809). After the meta-analysis, the number of genomic loci increased from 14 to 19 in ADHD, from 3 to 10 in ASD, from 45 to 57 in DEP, from 8 to 54 in BD, and from 64 to 87 in SCZ. We observed significant enrichment of overlapping genes among different disorders and identified a panel of cross-disorder genes. A total of seven genes were found being commonly associated with four out of five psychiatric conditions, namely GABBR1, GLT8D1, HIST1H1B, HIST1H2BN, HIST1H4L, KCNB1, and DCC. The SORCS3 gene was highlighted due to the fact that it was involved in all the five conditions of study. Analysis of correlations unveiled the existence of two clusters of related psychiatric conditions, SCZ and BD that were separate from the other three traits, and formed another group. Our results may provide a new insight for genetic basis of the five psychiatric disorders.

Identification of disease-associated loci using machine learning for genotype and network data integration

Motivation

Integration of different omics data could markedly help to identify biological signatures, understand the missing heritability of complex diseases and ultimately achieve personalized medicine. Standard regression models used in Genome-Wide Association Studies (GWAS) identify loci with a strong effect size, whereas GWAS meta-analyses are often needed to capture weak loci contributing to the missing heritability. Development of novel machine learning algorithms for merging genotype data with other omics data is highly needed as it could enhance the prioritization of weak loci.

Results

We developed cNMTF (corrected non-negative matrix tri-factorization), an integrative algorithm based on clustering techniques of biological data. This method assesses the inter-relatedness between genotypes, phenotypes, the damaging effect of the variants and gene networks in order to identify loci-trait associations. cNMTF was used to prioritize genes associated with lipid traits in two population cohorts. We replicated 129 genes reported in GWAS world-wide and provided evidence that supports 85% of our findings (226 out of 265 genes), including recent associations in literature (NLGN1), regulators of lipid metabolism (DAB1) and pleiotropic genes for lipid traits (CARM1). Moreover, cNMTF performed efficiently against strong population structures by accounting for the individuals’ ancestry. As the method is flexible in the incorporation of diverse omics data sources, it can be easily adapted to the user’s research needs.

Availability and implementation

An R package (cnmtf) is available at https://lgl15.github.io/cnmtf_web/index.html.

Supplementary information

Supplementary data are available at Bioinformatics online.

The Integrated Landscape of Biological Candidate Causal Genes in Coronary Artery Disease

Background

Genome-wide association studies (GWASs) have identified more than 150 genetic loci that demonstrate robust association with coronary artery disease (CAD). In contrast to the success of GWAS, the translation from statistical signals to biological mechanism and exploration of causal genes for drug development remain difficult, owing to the complexity of gene regulatory and linkage disequilibrium patterns. We aim to prioritize the plausible causal genes for CAD at a genome-wide level.

Methods

We integrated the latest GWAS summary statistics with other omics data from different layers and utilized eight different computational methods to predict CAD potential causal genes. The prioritized candidate genes were further characterized by pathway enrichment analysis, tissue-specific expression analysis, and pathway crosstalk analysis.

Results

Our analysis identified 55 high-confidence causal genes for CAD, among which 15 genes (LPL, COL4A2, PLG, CDKN2B, COL4A1, FES, FLT1, FN1, IL6R, LPA, PCSK9, PSRC1, SMAD3, SWAP70, and VAMP8) ranked the highest priority because of consistent evidence from different data-driven approaches. GO analysis showed that these plausible causal genes were enriched in lipid metabolic and extracellular regions. Tissue-specific enrichment analysis revealed that these genes were significantly overexpressed in adipose and liver tissues. Further, KEGG and crosstalk analysis also revealed several key pathways involved in the pathogenesis of CAD.

Conclusion

Our study delineated the landscape of CAD potential causal genes and highlighted several biological processes involved in CAD pathogenesis. Further studies and experimental validations of these genes may shed light on mechanistic insights into CAD development and provide potential drug targets for future therapeutics.

Enhancing reproducibility of gene expression analysis with known protein functional relationships: The concept of well-associated protein

Identification of differentially expressed genes (DEGs) is well recognized to be variable across independent replications of genome-wide transcriptional studies. These are often employed to characterize disease state early in the process of discovery and prioritize novel targets aimed at addressing unmet medical need. Increasing reproducibility of biological findings from these studies could potentially positively impact the success rate of new clinical interventions. This work demonstrates that statistically sound combination of gene expression data with prior knowledge about biology in the form of large protein interaction networks can yield quantitatively more reproducible observations from studies characterizing human disease. The novel concept of Well-Associated Proteins (WAPs) introduced herein-gene products significantly associated on protein interaction networks with the differences in transcript levels between control and disease-does not require choosing a differential expression threshold and can be computed efficiently enough to enable false discovery rate estimation via permutation. Reproducibility of WAPs is shown to be on average superior to that of DEGs under easily-quantifiable conditions suggesting that they can yield a significantly more robust description of disease. Enhanced reproducibility of WAPs versus DEGs is first demonstrated with four independent data sets focused on systemic sclerosis. This finding is then validated over thousands of pairs of data sets obtained by random partitions of large studies in several other diseases. Conditions that individual data sets must satisfy to yield robust WAP scores are examined. Reproducible identification of WAPs can potentially benefit drug target selection and precision medicine studies.

Synaptic and Gene Regulatory Mechanisms in Schizophrenia, Autism, and 22q11.2 Copy Number Variant-Mediated Risk for Neuropsychiatric Disorders

BACKGROUND

22q11.2 copy number variants are among the most highly penetrant genetic risk variants for developmental neuropsychiatric disorders such as schizophrenia (SCZ) and autism spectrum disorder (ASD). However, the specific mechanisms through which they confer risk remain unclear.

METHODS

Using a functional genomics approach, we integrated transcriptomic data from the developing human brain, genome-wide association findings for SCZ and ASD, protein interaction data, and gene expression signatures from SCZ and ASD postmortem cortex to 1) organize genes into the developmental cellular and molecular systems within which they operate, 2) identify neurodevelopmental processes associated with polygenic risk for SCZ and ASD across the allelic frequency spectrum, and 3) elucidate pathways and individual genes through which 22q11.2 copy number variants may confer risk for each disorder.

RESULTS

Polygenic risk for SCZ and ASD converged on partially overlapping neurodevelopmental modules involved in synaptic function and transcriptional regulation, with ASD risk variants additionally enriched for modules involved in neuronal differentiation during fetal development. The 22q11.2 locus formed a large protein network during development that disproportionately affected SCZ-associated and ASD-associated neurodevelopmental modules, including loading highly onto synaptic and gene regulatory pathways. SEPT5, PI4KA, and SNAP29 genes are candidate drivers of 22q11.2 synaptic pathology relevant to SCZ and ASD, and DGCR8 and HIRA are candidate drivers of disease-relevant alterations in gene regulation.

CONCLUSIONS

This approach offers a powerful framework to identify neurodevelopmental processes affected by diverse risk variants for SCZ and ASD and elucidate mechanisms through which highly penetrant, multigene copy number variants contribute to disease risk.

Convergent genomic signatures of high-altitude adaptation among domestic mammals

Abundant and diverse domestic mammals living on the Tibetan Plateau provide useful materials for investigating adaptive evolution and genetic convergence. Here, we used 327 genomes from horses, sheep, goats, cattle, pigs and dogs living at both high and low altitudes, including 73 genomes generated for this study, to disentangle the genetic mechanisms underlying local adaptation of domestic mammals. Although molecular convergence is comparatively rare at the DNA sequence level, we found convergent signature of positive selection at the gene level, particularly the EPAS1 gene in these Tibetan domestic mammals. We also reported a potential function in response to hypoxia for the gene C10orf67, which underwent positive selection in three of the domestic mammals. Our data provide an insight into adaptive evolution of high-altitude domestic mammals, and should facilitate the search for additional novel genes involved in the hypoxia response pathway.

A comparative study of the genetic components of three subcategories of autism spectrum disorder

The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) controversially combined previously distinct subcategories of autism spectrum disorder (ASD) into a single diagnostic category. However, genetic convergences and divergences between different ASD subcategories are unclear. By retrieving 1725 exonic de novo mutations (DNMs) from 1628 subjects with autistic disorder (AD), 1873 from 1564 subjects with pervasive developmental disorder not otherwise specified (PDD-NOS), 276 from 247 subjects with Asperger's syndrome (AS), and 2077 from 2299 controls, we found that rates of putative functional DNMs (loss-of-function, predicted deleterious missense, and frameshift) in all three subcategories were significantly higher than those in control. We then investigated the convergences and divergences of the three ASD subcategories based on four genetic aspects: whether any two ASD subcategories (1) shared significantly more genes with functional DNMs, (2) exhibited similar spatio-temporal expression patterns, (3) shared significantly more candidate genes, and (4) shared some ASD-associated functional pathways. It is revealed that AD and PDD-NOS were broadly convergent in terms of all four genetic aspects, suggesting these two ASD subcategories may be genetically combined. AS was divergent to AD and PDD-NOS for aspects of functional DNMs and expression patterns, whereas AS and AD/PDD-NOS were convergent for aspects of candidate genes and functional pathways. Our results indicated that the three ASD subcategories present more genetic convergences than divergences, favouring DSM-5's new classification. This study suggests that specifically defined genotypes and their corresponding phenotypes should be integrated analyzed for precise diagnosis of complex disorders, such as ASD.

DNA methylation profiles are associated with complex regional pain syndrome after traumatic injury

Factors contributing to development of complex regional pain syndrome (CRPS) are not fully understood. This study examined possible epigenetic mechanisms that may contribute to CRPS after traumatic injury. DNA methylation profiles were compared between individuals developing CRPS (n = 9) and those developing non-CRPS neuropathic pain (n = 38) after undergoing amputation following military trauma. Linear Models for Microarray (LIMMA) analyses revealed 48 differentially methylated cytosine-phosphate-guanine dinucleotide (CpG) sites between groups (unadjusted P's < 0.005), with the top gene COL11A1 meeting Bonferroni-adjusted P < 0.05. The second largest differential methylation was observed for the HLA-DRB6 gene, an immune-related gene linked previously to CRPS in a small gene expression study. For all but 7 of the significant CpG sites, the CRPS group was hypomethylated. Numerous functional Gene Ontology-Biological Process categories were significantly enriched (false discovery rate-adjusted q value <0.15), including multiple immune-related categories (eg, activation of immune response, immune system development, regulation of immune system processes, and antigen processing and presentation). Differentially methylated genes were more highly connected in human protein-protein networks than expected by chance (P < 0.05), supporting the biological relevance of the findings. Results were validated in an independent sample linking a DNA biobank with electronic health records (n = 126 CRPS phenotype, n = 19,768 non-CRPS chronic pain phenotype). Analyses using PrediXcan methodology indicated differences in the genetically determined component of gene expression in 7 of 48 genes identified in methylation analyses (P's < 0.02). Results suggest that immune- and inflammatory-related factors might confer risk of developing CRPS after traumatic injury. Validation findings demonstrate the potential of using electronic health records linked to DNA for genomic studies of CRPS.

Applications of molecular networks in biomedicine

Due to the large interdependence between the molecular components of living systems, many phenomena, including those related to pathologies, cannot be explained in terms of a single gene or a small number of genes. Molecular networks, representing different types of relationships between molecular entities, embody these large sets of interdependences in a framework that allow their mining from a systemic point of view to obtain information. These networks, often generated from high-throughput omics datasets, are used to study the complex phenomena of human pathologies from a systemic point of view. Complementing the reductionist approach of molecular biology, based on the detailed study of a small number of genes, systemic approaches to human diseases consider that these are better reflected in large and intricate networks of relationships between genes. These networks, and not the single genes, provide both better markers for diagnosing diseases and targets for treating them. Network approaches are being used to gain insight into the molecular basis of complex diseases and interpret the large datasets associated with them, such as genomic variants. Network formalism is also suitable for integrating large, heterogeneous and multilevel datasets associated with diseases from the molecular level to organismal and epidemiological scales. Many of these approaches are available to nonexpert users through standard software packages.

Inherited and De Novo Genetic Risk for Autism Impacts Shared Networks

We performed a comprehensive assessment of rare inherited variation in autism spectrum disorder (ASD) by analyzing whole-genome sequences of 2,308 individuals from families with multiple affected children. We implicate 69 genes in ASD risk, including 24 passing genome-wide Bonferroni correction and 16 new ASD risk genes, most supported by rare inherited variants, a substantial extension of previous findings. Biological pathways enriched for genes harboring inherited variants represent cytoskeletal organization and ion transport, which are distinct from pathways implicated in previous studies. Nevertheless, the de novo and inherited genes contribute to a common protein-protein interaction network. We also identified structural variants (SVs) affecting non-coding regions, implicating recurrent deletions in the promoters of DLG2 and NR3C2. Loss of nr3c2 function in zebrafish disrupts sleep and social function, overlapping with human ASD-related phenotypes. These data support the utility of studying multiplex families in ASD and are available through the Hartwell Autism Research and Technology portal.

Efficient pathway enrichment and network analysis of GWAS summary data using GSA-SNP2

Pathway-based analysis in genome-wide association study (GWAS) is being widely used to uncover novel multi-genic functional associations. Many of these pathway-based methods have been used to test the enrichment of the associated genes in the pathways, but exhibited low powers and were highly affected by free parameters. We present the novel method and software GSA-SNP2 for pathway enrichment analysis of GWAS P-value data. GSA-SNP2 provides high power, decent type I error control and fast computation by incorporating the random set model and SNP-count adjusted gene score. In a comparative study using simulated and real GWAS data, GSA-SNP2 exhibited high power and best prioritized gold standard positive pathways compared with six existing enrichment-based methods and two self-contained methods (alternative pathway analysis approach). Based on these results, the difference between pathway analysis approaches was investigated and the effects of the gene correlation structures on the pathway enrichment analysis were also discussed. In addition, GSA-SNP2 is able to visualize protein interaction networks within and across the significant pathways so that the user can prioritize the core subnetworks for further studies. GSA-SNP2 is freely available at https://sourceforge.net/projects/gsasnp2.

Interactome Analyses implicated CAMK2A in the genetic predisposition and pharmacological mechanism of Bipolar Disorder

Bipolar disorder (BPD) is a severe mental illness characterized by fluctuations in mood states, behaviors and energy levels. Growing evidence suggests that genes associated with specific illnesses tend to interact together and encode a tight protein-protein interaction (PPI) network, providing valuable information for understanding their pathogenesis. To gain insights into the genetic and physiological foundation of BPD, we conduct the physical PPI analysis of 184 BPD risk genes distilled from genome-wide association studies and exome sequencing studies. We have identified several hub genes (CAMK2A, HSP90AA1 and PLCG1) among those risk genes, and observed significant enrichment of the BPD risk genes in certain pathways such as calcium signaling, oxytocin signaling and circadian entrainment. Furthermore, while none of the 184 genetic risk genes are "well established" BPD drug targets, our PPI analysis showed that αCaMKII (encoded by CAMK2A) had direct physical PPIs with targets (HRH1, SCN5A and CACNA1E) of clinically used anti-manic BPD drugs, such as carbamazepine. We thus speculated that αCaMKII might be involved in the cellular pharmacological actions of those drugs. Using cultured rat primary cortical neurons, we found that carbamazepine treatment induced phosphorylation of αCaMKII in dose-dependent manners. Intriguingly, previous study showed that CAMK2A heterozygous knockout (CAMK2A^+/-) mice exhibited infradian oscillation of locomotor activities that can be rescued by carbamazepine. Our data, in combination with previous studies, provide convergent evidence for the involvement of CAMK2A in the risk of BPD.