Landscape of allele-specific transcription factor binding in the human genome

Perspectives on Codebook: sequence specificity of uncharacterized human transcription factors

We describe an effort ("Codebook") to determine the sequence specificity of 332 putative and largely uncharacterized human transcription factors (TFs), as well as 61 control TFs. Nearly 5,000 independent experiments across multiple in vitro and in vivo assays produced motifs for just over half of the putative TFs analyzed (177, or 53%), of which most are unique to a single TF. The data highlight the extensive contribution of transposable elements to TF evolution, both in cis and trans, and identify tens of thousands of conserved, base-level binding sites in the human genome. The use of multiple assays provides an unprecedented opportunity to benchmark and analyze TF sequence specificity, function, and evolution, as further explored in accompanying manuscripts. 1,421 human TFs are now associated with a DNA binding motif. Extrapolation from the Codebook benchmarking, however, suggests that many of the currently known binding motifs for well-studied TFs may inaccurately describe the TF's true sequence preferences.

Disease-Linked Regulatory DNA Variants and Homeostatic Transcription Factors in Epidermis

Identifying noncoding single nucleotide variants ( SNVs ) in regulatory DNA linked to polygenic disease risk, the transcription factors ( TFs ) they bind, and the target genes they dysregulate is a goal in polygenic disease research. Massively parallel reporter gene analysis ( MPRA ) of 3,451 SNVs linked to risk for polygenic skin diseases characterized by disrupted epidermal homeostasis identified 355 differentially active SNVs ( daSNVs ). daSNV target gene analysis, combined with daSNV editing, underscored dysregulated epidermal differentiation as a pathomechanism shared across common polygenic skin diseases. CRISPR knockout screens of 1772 human TFs revealed 108 TFs essential for epidermal progenitor differentiation, uncovering novel roles for ZNF217, CXXC1, FOXJ2, IRX2 and NRF1. Population sampling CUT&RUN of 27 homeostatic TFs identified allele-specific DNA binding ( ASB ) differences at daSNVs enriched near epidermal homeostasis and monogenic skin disease genes, with notable representation of SP/KLF and AP-1/2 TFs. This resource implicates dysregulated differentiation in risk for diverse polygenic skin diseases.

Benchmarking and building DNA binding affinity models using allele-specific and allele-agnostic transcription factor binding data

BACKGROUND

Transcription factors (TFs) bind to DNA in a highly sequence-specific manner. This specificity manifests itself in vivo as differences in TF occupancy between the two alleles at heterozygous loci. Genome-scale assays such as ChIP-seq currently are limited in their power to detect allele-specific binding (ASB) both in terms of read coverage and representation of individual variants in the cell lines used. This makes prediction of allelic differences in TF binding from sequence alone desirable, provided that the reliability of such predictions can be quantitatively assessed.

RESULTS

We here propose methods for benchmarking sequence-to-affinity models for TF binding in terms of their ability to predict allelic imbalances in ChIP-seq counts. We use a likelihood function based on an over-dispersed binomial distribution to aggregate evidence for allelic preference across the genome without requiring statistical significance for individual variants. This allows us to systematically compare predictive performance when multiple binding models for the same TF are available. To facilitate the de novo inference of high-quality models from paired-end in vivo binding data such as ChIP-seq, ChIP-exo, and CUT&Tag without read mapping or peak calling, we introduce an extensible reimplementation of our biophysically interpretable machine learning framework named PyProBound. Explicitly accounting for assay-specific bias in DNA fragmentation rate when training on ChIP-seq yields improved TF binding models. Moreover, we show how PyProBound can leverage our threshold-free ASB likelihood function to perform de novo motif discovery using allele-specific ChIP-seq counts.

CONCLUSION

Our work provides new strategies for predicting the functional impact of non-coding variants.

Characterization of non-coding variants associated with transcription factor binding through ATAC-seq-defined footprint QTLs in liver

Non-coding variants discovered by genome-wide association studies (GWAS) are enriched in regulatory elements harboring transcription factor (TF) binding motifs, strongly suggesting a connection between disease association and the disruption of cis-regulatory sequences. Occupancy of a TF inside a region of open chromatin can be detected in ATAC-seq where bound TFs block the transposase Tn5, leaving a pattern of relatively depleted Tn5 insertions known as a "footprint". Here, we sought to identify variants associated with TF-binding, or "footprint quantitative trait loci" (fpQTLs) in ATAC-seq data generated from 170 human liver samples. We used computational tools to scan the ATAC-seq reads to quantify TF binding likelihood as "footprint scores" at variants derived from whole genome sequencing generated in the same samples. We tested for association between genotype and footprint score and observed 693 fpQTLs associated with footprint-inferred TF binding (FDR < 5%). Given that Tn5 insertion sites are measured with base-pair resolution, we show that fpQTLs can aid GWAS and QTL fine-mapping by precisely pinpointing TF activity within broad trait-associated loci where the underlying causal variant is unknown. Liver fpQTLs were strongly enriched across ChIP-seq peaks, liver expression QTLs (eQTLs), and liver-related GWAS loci, and their inferred effect on TF binding was concordant with their effect on underlying sequence motifs in 80% of cases. We conclude that fpQTLs can reveal causal GWAS variants, define the role of TF binding site disruption in disease and provide functional insights into non-coding variants, ultimately informing novel treatments for common diseases.

Autoimmunity-Associated SNP rs3024505 Disrupts STAT3 Binding in B Cells, Leading to IL10 Dysregulation

Interleukin 10 (IL10) is a major anti-inflammatory cytokine that acts as a master regulator of the immune response. A single nucleotide polymorphism rs3024505(C/T), located downstream of the IL10 gene, is associated with several aggressive inflammatory diseases, including systemic lupus erythematosus, Sjögren's syndrome, Crohn's disease, and ulcerative colitis. In such autoimmune pathologies, IL10-producing B cells play a protective role by decreasing the level of inflammation and restoring immune homeostasis. This study demonstrates that rs3024505 is located within an enhancer that augments the activity of the IL10 promoter in a reporter system based on a human B cell line. The common rs3024505(C) variant creates a functional binding site for the transcription factor STAT3, whereas the risk allele rs3024505(T) disrupts STAT3 binding, thereby reducing the IL10 promoter activity. Our findings indicate that B cells from individuals carrying the minor rs3024505(T) allele may produce less IL10 due to the disrupted STAT3 binding site, contributing to the progression of inflammatory pathologies.

Semi-supervised learning with pseudo-labeling compares favorably with large language models for regulatory sequence prediction

Predicting molecular processes using deep learning is a promising approach to provide biological insights for non-coding single nucleotide polymorphisms identified in genome-wide association studies. However, most deep learning methods rely on supervised learning, which requires DNA sequences associated with functional data, and whose amount is severely limited by the finite size of the human genome. Conversely, the amount of mammalian DNA sequences is growing exponentially due to ongoing large-scale sequencing projects, but in most cases without functional data. To alleviate the limitations of supervised learning, we propose a novel semi-supervised learning (SSL) based on pseudo-labeling, which allows to exploit unlabeled DNA sequences from numerous genomes during model pre-training. We further improved it incorporating principles from the Noisy Student algorithm to predict the confidence in pseudo-labeled data used for pre-training, which showed improvements for transcription factor with very few binding (very small training data). The approach is very flexible and can be used to train any neural architecture including state-of-the-art models, and shows in most cases strong predictive performance improvements compared to standard supervised learning. Moreover, small models trained by SSL showed similar or better performance than large language model DNABERT2.

Allele-specific transcription factor binding across human brain regions offers mechanistic insight into eQTLs

Transcription factors (TFs) regulate gene expression by facilitating or disrupting the formation of transcription initiation machinery at particular genomic loci. Because TF occupancy is driven in part by recognition of DNA sequence, genetic variation can influence TF-DNA associations and gene regulation. To identify variants that impact TF binding in human brain tissues, we assessed allele-specific binding (ASB) at heterozygous variants for 94 TFs in nine brain regions from two donors. Leveraging graph genomes constructed from phased genomic sequence data, we compared ChIP-seq signals between alleles at heterozygous variants within each brain region and identified thousands of variants exhibiting ASB for at least one TF. ASB reproducibility was measured by comparisons between independent experiments both within and between donors. We found that rare alleles in the general population more frequently led to reduced TF binding, whereas common alleles had an equal likelihood of increasing or decreasing binding. Further, for ASB variants in predicted binding motifs, the favored allele tended to be the one with the stronger expected motif match, but this concordance was not observed within highly occupied sites. We also found that neuron-specific cis-regulatory elements (cCREs), in contrast with oligodendrocyte-specific cCREs, showed depletion of ASB variants. We identified 2670 ASB variants associated with evidence for allele-specific gene expression in the brain from GTEx data and observed increasing eQTL effect direction concordance as ASB significance increases. These results provide a valuable and unique resource for mechanistic analysis of cis-regulatory variation in human brain tissue.

Identifying genetic variants associated with chromatin looping and genome function

Here we present a comprehensive HiChIP dataset on naïve CD4 T cells (nCD4) from 30 donors and identify QTLs that associate with genotype-dependent and/or allele-specific variation of HiChIP contacts defining loops between active regulatory regions (iQTLs). We observe a substantial overlap between iQTLs and previously defined eQTLs and histone QTLs, and an enrichment for fine-mapped QTLs and GWAS variants. Furthermore, we describe a distinct subset of nCD4 iQTLs, for which the significant variation of chromatin contacts in nCD4 are translated into significant eQTL trends in CD4 T cell memory subsets. Finally, we define connectivity-QTLs as iQTLs that are significantly associated with concordant genotype-dependent changes in chromatin contacts over a broad genomic region (e.g., GWAS SNP in the RNASET2 locus). Our results demonstrate the importance of chromatin contacts as a complementary modality for QTL mapping and their power in identifying previously uncharacterized QTLs linked to cell-specific gene expression and connectivity.

Revealing novel genomic insights and therapeutic targets for juvenile idiopathic arthritis through omics

BACKGROUND

The genetic architecture of JIA remains only partially comprehended. There is a clear imperative for continued endeavours to uncover insights into the underlying causes of JIA.

METHODS

This study encompassed a comprehensive spectrum of endeavours, including conducting a JIA genome-wide association study (GWAS) meta-analysis that incorporated data from 4550 JIA cases and 18 446 controls. We employed in silico and genome-editing approaches to prioritizing target genes. To investigate pleiotropic effects, we conducted phenome-wide association studies. Cell-type enrichment analyses were performed by integrating bulk and single-cell sequencing data. Finally, we delved into potential druggable targets for JIA.

RESULTS

Fourteen genome-wide significant non-HLA loci were identified, including four novel loci, each exhibiting pleiotropic associations with other autoimmune diseases or musculoskeletal traits. We uncovered strong genetic correlation between JIA and BMD traits at 52 genomic regions, including three GWAS loci for JIA. Candidate genes with immune functions were captured by in silico analyses at each novel locus, with additional findings identified through our experimental approach. Cell-type enrichment analysis revealed 21 specific immune cell types crucial for the affected organs in JIA, indicating their potential contribution to the disease. Finally, 24 known or candidate druggable target genes were prioritized.

CONCLUSIONS

Our identification of four novel JIA-associated genes, CD247, RHOH, COLEC10 and IRF8, broadens the novel potential drug repositioning opportunities. We established a new genetic link between COLEC10, TNFRSF11B and JIA/BMD. Additionally, the identification of RHOH underscores its role in positive thymocyte selection, thereby illuminating a critical facet of JIA's underlying biological mechanisms.

Multi-Omics Analysis Revealed the rSNPs Potentially Involved in T2DM Pathogenic Mechanism and Metformin Response

The goal of our study was to identify and assess the functionally significant SNPs with potentially important roles in the development of type 2 diabetes mellitus (T2DM) and/or their effect on individual response to antihyperglycemic medication with metformin. We applied a bioinformatics approach to identify the regulatory SNPs (rSNPs) associated with allele-asymmetric binding and expression events in our paired ChIP-seq and RNA-seq data for peripheral blood mononuclear cells (PBMCs) of nine healthy individuals. The rSNP outcomes were analyzed using public data from the GWAS (Genome-Wide Association Studies) and Genotype-Tissue Expression (GTEx). The differentially expressed genes (DEGs) between healthy and T2DM individuals (GSE221521), including metformin responders and non-responders (GSE153315), were searched for in GEO RNA-seq data. The DEGs harboring rSNPs were analyzed using the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). We identified 14,796 rSNPs in the promoters of 5132 genes of human PBMCs. We found 4280 rSNPs to associate with both phenotypic traits (GWAS) and expression quantitative trait loci (eQTLs) from GTEx. Between T2DM patients and controls, 3810 rSNPs were detected in the promoters of 1284 DEGs. Based on the protein-protein interaction (PPI) network, we identified 31 upregulated hub genes, including the genes involved in inflammation, obesity, and insulin resistance. The top-ranked 10 enriched KEGG pathways for these hubs included insulin, AMPK, and FoxO signaling pathways. Between metformin responders and non-responders, 367 rSNPs were found in the promoters of 131 DEGs. Genes encoding transcription factors and transcription regulators were the most widely represented group and many were shown to be involved in the T2DM pathogenesis. We have formed a list of human rSNPs that add functional interpretation to the T2DM-association signals identified in GWAS. The results suggest candidate causal regulatory variants for T2DM, with strong enrichment in the pathways related to glucose metabolism, inflammation, and the effects of metformin.

DNA-binding factor footprints and enhancer RNAs identify functional non-coding genetic variants

BACKGROUND

Genome-wide association studies (GWAS) have revealed a multitude of candidate genetic variants affecting the risk of developing complex traits and diseases. However, the highlighted regions are typically in the non-coding genome, and uncovering the functional causative single nucleotide variants (SNVs) is challenging. Prioritization of variants is commonly based on genomic annotation with markers of active regulatory elements, but current approaches still poorly predict functional variants. To address this, we systematically analyze six markers of active regulatory elements for their ability to identify functional variants.

RESULTS

We benchmark against molecular quantitative trait loci (molQTL) from assays of regulatory element activity that identify allelic effects on DNA-binding factor occupancy, reporter assay expression, and chromatin accessibility. We identify the combination of DNase footprints and divergent enhancer RNA (eRNA) as markers for functional variants. This signature provides high precision, but with a trade-off of low recall, thus substantially reducing candidate variant sets to prioritize variants for functional validation. We present this as a framework called FINDER-Functional SNV IdeNtification using DNase footprints and eRNA.

CONCLUSIONS

We demonstrate the utility to prioritize variants using leukocyte count trait and analyze variants in linkage disequilibrium with a lead variant to predict a functional variant in asthma. Our findings have implications for prioritizing variants from GWAS, in development of predictive scoring algorithms, and for functionally informed fine mapping approaches.

Allele-specific binding variants causing ChIP-seq peak height of histone modification are not enriched in expression QTL annotations

BACKGROUND

Genome sequence variants affecting complex traits (quantitative trait loci, QTL) are enriched in functional regions of the genome, such as those marked by certain histone modifications. These variants are believed to influence gene expression. However, due to the linkage disequilibrium among nearby variants, pinpointing the precise location of QTL is challenging. We aimed to identify allele-specific binding (ASB) QTL (asbQTL) that cause variation in the level of histone modification, as measured by the height of peaks assayed by ChIP-seq (chromatin immunoprecipitation sequencing). We identified DNA sequences that predict the difference between alleles in ChIP-seq peak height in H3K4me3 and H3K27ac histone modifications in the mammary glands of cows.

RESULTS

We used a gapped k-mer support vector machine, a novel best linear unbiased prediction model, and a multiple linear regression model that combines the other two approaches to predict variant impacts on peak height. For each method, a subset of 1000 sites with the highest magnitude of predicted ASB was considered as candidate asbQTL. The accuracy of this prediction was measured by the proportion where the predicted direction matched the observed direction. Prediction accuracy ranged between 0.59 and 0.74, suggesting that these 1000 sites are enriched for asbQTL. Using independent data, we investigated functional enrichment in the candidate asbQTL set and three control groups, including non-causal ASB sites, non-ASB variants under a peak, and SNPs (single nucleotide polymorphisms) not under a peak. For H3K4me3, a higher proportion of the candidate asbQTL were confirmed as ASB when compared to the non-causal ASB sites (P < 0.01). However, these candidate asbQTL did not enrich for the other annotations, including expression QTL (eQTL), allele-specific expression QTL (aseQTL) and sites conserved across mammals (P > 0.05).

CONCLUSIONS

We identified putatively causal sites for asbQTL using the DNA sequence surrounding these sites. Our results suggest that many sites influencing histone modifications may not directly affect gene expression. However, it is important to acknowledge that distinguishing between putative causal ASB sites and other non-causal ASB sites in high linkage disequilibrium with the causal sites regarding their impact on gene expression may be challenging due to limitations in statistical power.

Identify Regulatory eQTLs by Multiome Sequencing in Prostate Single Cells

While genome-wide association studies and expression quantitative trait loci (eQTL) analysis have made significant progress in identifying noncoding variants associated with prostate cancer risk and bulk tissue transcriptome changes, the regulatory effect of these genetic elements on gene expression remains largely unknown. Recent developments in single-cell sequencing have made it possible to perform ATAC-seq and RNA-seq profiling simultaneously to capture functional associations between chromatin accessibility and gene expression. In this study, we tested our hypothesis that this multiome single-cell approach allows for mapping regulatory elements and their target genes at prostate cancer risk loci. We applied a 10X Multiome ATAC + Gene Expression platform to encapsulate Tn5 transposase-tagged nuclei from multiple prostate cell lines for a total of 65,501 high quality single cells from RWPE1, RWPE2, PrEC, BPH1, DU145, PC3, 22Rv1 and LNCaP cell lines. To address data sparsity commonly seen in the single-cell sequencing, we performed targeted sequencing to enrich sequencing data at prostate cancer risk loci involving 2,730 candidate germline variants and 273 associated genes. Although not increasing the number of captured cells, the targeted multiome data did improve eQTL gene expression abundance by about 20% and chromatin accessibility abundance by about 5%. Based on this multiomic profiling, we further associated RNA expression alterations with chromatin accessibility of germline variants at single cell levels. Cross validation analysis showed high overlaps between the multiome associations and the bulk eQTL findings from GTEx prostate cohort. We found that about 20% of GTEx eQTLs were covered within the significant multiome associations (p-value ≤ 0.05, gene abundance percentage ≥ 5%), and roughly 10% of the multiome associations could be identified by significant GTEx eQTLs. We also analyzed accessible regions with available heterozygous SNP reads and observed more frequent association in genomic regions with allelically accessible variants (p = 0.0055). Among these findings were previously reported regulatory variants including rs60464856-RUVBL1 (multiome p-value = 0.0099 in BPH1) and rs7247241-SPINT2 (multiome p-value = 0.0002- 0.0004 in 22Rv1). We also functionally validated a new regulatory SNP and its target gene rs2474694-VPS53 (multiome p-value = 0.00956 in BPH1 and 0.00625 in DU145) by reporter assay and SILAC proteomics sequencing. Taken together, our data demonstrated the feasibility of the multiome single-cell approach for identifying regulatory SNPs and their regulated genes.

Methods for Functional Characterization of Genetic Polymorphisms of Non-Coding Regulatory Regions of the Human Genome

Currently, numerous associations between genetic polymorphisms and various diseases have been characterized through the Genome-Wide Association Studies. Majority of the clinically significant polymorphisms are localized in non-coding regions of the genome. While modern bioinformatic resources make it possible to predict molecular mechanisms that explain influence of the non-coding polymorphisms on gene expression, such hypotheses require experimental verification. This review discusses the methods for elucidating molecular mechanisms underlying dependence of the disease pathogenesis on specific genetic variants within the non-coding sequences. A particular focus is on the methods for identification of transcription factors with binding efficiency dependent on polymorphic variations. Despite remarkable progress in bioinformatic resources enabling prediction of the impact of polymorphisms on the disease pathogenesis, there is still the need for experimental approaches to investigate this issue.

Functional dissection of complex and molecular trait variants at single nucleotide resolution

Identifying the causal variants and mechanisms that drive complex traits and diseases remains a core problem in human genetics. The majority of these variants have individually weak effects and lie in non-coding gene-regulatory elements where we lack a complete understanding of how single nucleotide alterations modulate transcriptional processes to affect human phenotypes. To address this, we measured the activity of 221,412 trait-associated variants that had been statistically fine-mapped using a Massively Parallel Reporter Assay (MPRA) in 5 diverse cell-types. We show that MPRA is able to discriminate between likely causal variants and controls, identifying 12,025 regulatory variants with high precision. Although the effects of these variants largely agree with orthogonal measures of function, only 69% can plausibly be explained by the disruption of a known transcription factor (TF) binding motif. We dissect the mechanisms of 136 variants using saturation mutagenesis and assign impacted TFs for 91% of variants without a clear canonical mechanism. Finally, we provide evidence that epistasis is prevalent for variants in close proximity and identify multiple functional variants on the same haplotype at a small, but important, subset of trait-associated loci. Overall, our study provides a systematic functional characterization of likely causal common variants underlying complex and molecular human traits, enabling new insights into the regulatory grammar underlying disease risk.

Non-coding DNA variants for risk in lupus

Systemic Lupus Erythematosus (SLE) is a multifactorial autoimmune disease that arises from a dynamic interplay between genetics and environmental triggers. The advent of sophisticated genomics technology has catalyzed a shift in our understanding of disease etiology, spotlighting the pivotal role of non-coding DNA variants in SLE pathogenesis. In this review, we present a comprehensive examination of the non-coding variants associated with SLE, shedding light on their role in influencing disease risk and progression. We discuss the latest methodological advancements that have been instrumental in the identification and functional characterization of these genomic elements, with a special focus on the transformative power of CRISPR-based gene-editing technologies. Additionally, the review probes into the therapeutic opportunities that arise from modulating non-coding regions associated with SLE. Through an exploration of the complex network of non-coding DNA, this review aspires to decode the genetic puzzle of SLE and set the stage for groundbreaking gene-based therapeutic interventions and the advancement of precision medicine strategies tailored to SLE management.

Shared genomic segments analysis identifies MHC class I and class III molecules as genetic risk factors for juvenile idiopathic arthritis

Juvenile idiopathic arthritis (JIA) is a complex rheumatic disease encompassing several clinically defined subtypes of varying severity. The etiology of JIA remains largely unknown, but genome-wide association studies (GWASs) have identified up to 22 genes associated with JIA susceptibility, including a well-established association with HLA-DRB1. Continued investigation of heritable risk factors has been hindered by disease heterogeneity and low disease prevalence. In this study, we utilized shared genomic segments (SGS) analysis on whole-genome sequencing of 40 cases from 12 multi-generational pedigrees significantly enriched for JIA. Subsets of cases are connected by a common ancestor in large extended pedigrees, increasing the power to identify disease-associated loci. SGS analysis identifies genomic segments shared among disease cases that are likely identical by descent and anchored by a disease locus. This approach revealed statistically significant signals for major histocompatibility complex (MHC) class I and class III alleles, particularly HLA-A∗02:01, which was observed at a high frequency among cases. Furthermore, we identified an additional risk locus at 12q23.2-23.3, containing genes primarily expressed by naive B cells, natural killer cells, and monocytes. The recognition of additional risk beyond HLA-DRB1 provides a new perspective on immune cell dynamics in JIA. These findings contribute to our understanding of JIA and may guide future research and therapeutic strategies.

rs822336 binding to C/EBPβ and NFIC modulates induction of PD-L1 expression and predicts anti-PD-1/PD-L1 therapy in advanced NSCLC

Efficient predictive biomarkers are needed for immune checkpoint inhibitor (ICI)-based immunotherapy in non-small cell lung cancer (NSCLC). Testing the predictive value of single nucleotide polymorphisms (SNPs) in programmed cell death 1 (PD-1) or its ligand 1 (PD-L1) has shown contrasting results. Here, we aim to validate the predictive value of PD-L1 SNPs in advanced NSCLC patients treated with ICIs as well as to define the molecular mechanisms underlying the role of the identified SNP candidate. rs822336 efficiently predicted response to anti-PD-1/PD-L1 immunotherapy in advanced non-oncogene addicted NSCLC patients as compared to rs2282055 and rs4143815. rs822336 mapped to the promoter/enhancer region of PD-L1, differentially affecting the induction of PD-L1 expression in human NSCLC cell lines as well as their susceptibility to HLA class I antigen matched PBMCs incubated with anti-PD-1 monoclonal antibody nivolumab. The induction of PD-L1 expression by rs822336 was mediated by a competitive allele-specificity binding of two identified transcription factors: C/EBPβ and NFIC. As a result, silencing of C/EBPβ and NFIC differentially regulated the induction of PD-L1 expression in human NSCLC cell lines carrying different rs822336 genotypes. Analysis by binding microarray further validated the competitive allele-specificity binding of C/EBPβ and NFIC to PD-L1 promoter/enhancer region based on rs822336 genotype in human NSCLC cell lines. These findings have high clinical relevance since identify rs822336 and induction of PD-L1 expression as novel biomarkers for predicting anti-PD-1/PD-L1-based immunotherapy in advanced NSCLC patients.

HaplotagLR: An efficient and configurable utility for haplotagging long reads

Understanding the functional effects of sequence variation is crucial in genomics. Individual human genomes contain millions of variants that contribute to phenotypic variability and disease risks at the population level. Because variants rarely act in isolation, we must consider potential interactions of neighboring variants to accurately predict functional effects. We can accomplish this using haplotagging, which matches sequencing reads to their parental haplotypes using alleles observed at known heterozygous variants. However, few published tools for haplotagging exist and these share several technical and usability-related shortcomings that limit applicability, in particular a lack of insight or control over error rates, and lack of key metrics on the underlying sources of haplotagging error. Here we present HaplotagLR: a user-friendly tool that haplotags long sequencing reads based on a multinomial model and existing phased variant lists. HaplotagLR is user-configurable and includes a basic error model to control the empirical FDR in its output. We show that HaplotagLR outperforms the leading haplotagging method in simulated datasets, especially at high levels of specificity, and displays 7% greater sensitivity in haplotagging real data. HaplotagLR advances both the immediate utility of haplotagging and paves the way for further improvements to this important method.

AStruct: detection of allele-specific RNA secondary structure in structuromic probing data

BACKGROUND

Uncovering functional genetic variants from an allele-specific perspective is of paramount importance in advancing our understanding of gene regulation and genetic diseases. Recently, various allele-specific events, such as allele-specific gene expression, allele-specific methylation, and allele-specific binding, have been explored on a genome-wide scale due to the development of high-throughput sequencing methods. RNA secondary structure, which plays a crucial role in multiple RNA-associated processes like RNA modification, translation and splicing, has emerged as an essential focus of relevant research. However, tools to identify genetic variants associated with allele-specific RNA secondary structures are still lacking.

RESULTS

Here, we develop a computational tool called 'AStruct' that enables us to detect allele-specific RNA secondary structure (ASRS) from RT-stop based structuromic probing data. AStruct shows robust performance in both simulated datasets and public icSHAPE datasets. We reveal that single nucleotide polymorphisms (SNPs) with higher AStruct scores are enriched in coding regions and tend to be functional. These SNPs are highly conservative, have the potential to disrupt sites involved in m6A modification or protein binding, and are frequently associated with disease.

CONCLUSIONS

AStruct is a tool dedicated to invoke allele-specific RNA secondary structure events at heterozygous SNPs in RT-stop based structuromic probing data. It utilizes allelic variants, base pairing and RT-stop information under different cell conditions to detect dynamic and functional ASRS. Compared to sequence-based tools, AStruct considers dynamic cell conditions and outperforms in detecting functional variants. AStruct is implemented in JAVA and is freely accessible at: https://github.com/canceromics/AStruct .

Complex regulatory networks influence pluripotent cell state transitions in human iPSCs

Stem cells exist in vitro in a spectrum of interconvertible pluripotent states. Analyzing hundreds of hiPSCs derived from different individuals, we show the proportions of these pluripotent states vary considerably across lines. We discover 13 gene network modules (GNMs) and 13 regulatory network modules (RNMs), which are highly correlated with each other suggesting that the coordinated co-accessibility of regulatory elements in the RNMs likely underlie the coordinated expression of genes in the GNMs. Epigenetic analyses reveal that regulatory networks underlying self-renewal and pluripotency are more complex than previously realized. Genetic analyses identify thousands of regulatory variants that overlapped predicted transcription factor binding sites and are associated with chromatin accessibility in the hiPSCs. We show that the master regulator of pluripotency, the NANOG-OCT4 Complex, and its associated network are significantly enriched for regulatory variants with large effects, suggesting that they play a role in the varying cellular proportions of pluripotency states between hiPSCs. Our work bins tens of thousands of regulatory elements in hiPSCs into discrete regulatory networks, shows that pluripotency and self-renewal processes have a surprising level of regulatory complexity, and suggests that genetic factors may contribute to cell state transitions in human iPSC lines.

Systematic identification of genotype-dependent enhancer variants in eosinophilic esophagitis

Eosinophilic esophagitis (EoE) is a rare atopic disorder associated with esophageal dysfunction, including difficulty swallowing, food impaction, and inflammation, that develops in a small subset of people with food allergies. Genome-wide association studies (GWASs) have identified 9 independent EoE risk loci reaching genome-wide significance (p < 5 × 10-8) and 27 additional loci of suggestive significance (5 × 10-8 < p < 1 × 10-5). In the current study, we perform linkage disequilibrium (LD) expansion of these loci to nominate a set of 531 variants that are potentially causal. To systematically interrogate the gene regulatory activity of these variants, we designed a massively parallel reporter assay (MPRA) containing the alleles of each variant within their genomic sequence context cloned into a GFP reporter library. Analysis of reporter gene expression in TE-7, HaCaT, and Jurkat cells revealed cell-type-specific gene regulation. We identify 32 allelic enhancer variants, representing 6 genome-wide significant EoE loci and 7 suggestive EoE loci, that regulate reporter gene expression in a genotype-dependent manner in at least one cellular context. By annotating these variants with expression quantitative trait loci (eQTL) and chromatin looping data in related tissues and cell types, we identify putative target genes affected by genetic variation in individuals with EoE. Transcription factor enrichment analyses reveal possible roles for cell-type-specific regulators, including GATA3. Our approach reduces the large set of EoE-associated variants to a set of 32 with allelic regulatory activity, providing functional insights into the effects of genetic variation in this disease.

Comparative analysis of models in predicting the effects of SNPs on TF-DNA binding using large-scale in vitro and in vivo data

Non-coding variants associated with complex traits can alter the motifs of transcription factor (TF)-deoxyribonucleic acid binding. Although many computational models have been developed to predict the effects of non-coding variants on TF binding, their predictive power lacks systematic evaluation. Here we have evaluated 14 different models built on position weight matrices (PWMs), support vector machines, ordinary least squares and deep neural networks (DNNs), using large-scale in vitro (i.e. SNP-SELEX) and in vivo (i.e. allele-specific binding, ASB) TF binding data. Our results show that the accuracy of each model in predicting SNP effects in vitro significantly exceeds that achieved in vivo. For in vitro variant impact prediction, kmer/gkm-based machine learning methods (deltaSVM_HT-SELEX, QBiC-Pred) trained on in vitro datasets exhibit the best performance. For in vivo ASB variant prediction, DNN-based multitask models (DeepSEA, Sei, Enformer) trained on the ChIP-seq dataset exhibit relatively superior performance. Among the PWM-based methods, tRap demonstrates better performance in both in vitro and in vivo evaluations. In addition, we find that TF classes such as basic leucine zipper factors could be predicted more accurately, whereas those such as C2H2 zinc finger factors are predicted less accurately, aligning with the evolutionary conservation of these TF classes. We also underscore the significance of non-sequence factors such as cis-regulatory element type, TF expression, interactions and post-translational modifications in influencing the in vivo predictive performance of TFs. Our research provides valuable insights into selecting prioritization methods for non-coding variants and further optimizing such models.

HOCOMOCO in 2024: a rebuild of the curated collection of binding models for human and mouse transcription factors

We present a major update of the HOCOMOCO collection that provides DNA binding specificity patterns of 949 human transcription factors and 720 mouse orthologs. To make this release, we performed motif discovery in peak sets that originated from 14 183 ChIP-Seq experiments and reads from 2554 HT-SELEX experiments yielding more than 400 thousand candidate motifs. The candidate motifs were annotated according to their similarity to known motifs and the hierarchy of DNA-binding domains of the respective transcription factors. Next, the motifs underwent human expert curation to stratify distinct motif subtypes and remove non-informative patterns and common artifacts. Finally, the curated subset of 100 thousand motifs was supplied to the automated benchmarking to select the best-performing motifs for each transcription factor. The resulting HOCOMOCO v12 core collection contains 1443 verified position weight matrices, including distinct subtypes of DNA binding motifs for particular transcription factors. In addition to the core collection, HOCOMOCO v12 provides motif sets optimized for the recognition of binding sites in vivo and in vitro, and for annotation of regulatory sequence variants. HOCOMOCO is available at https://hocomoco12.autosome.org and https://hocomoco.autosome.org.

Role of Genetic Variation in Transcriptional Regulatory Elements in Heart Rhythm

Genetic predisposition to cardiac arrhythmias has been a field of intense investigation. Research initially focused on rare hereditary arrhythmias, but over the last two decades, the role of genetic variation (single nucleotide polymorphisms) in heart rate, rhythm, and arrhythmias has been taken into consideration as well. In particular, genome-wide association studies have identified hundreds of genomic loci associated with quantitative electrocardiographic traits, atrial fibrillation, and less common arrhythmias such as Brugada syndrome. A significant number of associated variants have been found to systematically localize in non-coding regulatory elements that control the tissue-specific and temporal transcription of genes encoding transcription factors, ion channels, and other proteins. However, the identification of causal variants and the mechanism underlying their impact on phenotype has proven difficult due to the complex tissue-specific, time-resolved, condition-dependent, and combinatorial function of regulatory elements, as well as their modest conservation across different model species. In this review, we discuss research efforts aimed at identifying and characterizing-trait-associated variant regulatory elements and the molecular mechanisms underlying their impact on heart rate or rhythm.

Benchmarking DNA binding affinity models using allele-specific transcription factor binding data

Transcription factors (TFs) bind to DNA in a highly sequence-specific manner. This specificity can manifest itself in vivo at heterozygous loci as a difference in TF occupancy between the two alleles. When applied on a genomic scale, functional genomic assays such as ChIP-seq typically lack the statistical power to detect allele-specific binding (ASB) at the level of individual variants. To address this, we propose a framework for benchmarking sequence-to-affinity models for TF binding in terms of their ability to predict allelic imbalances in ChIP-seq counts. We show that a likelihood function based on an over-dispersed binomial distribution can aggregate evidence for allelic preference across the genome without requiring statistical significance for individual variants. This allows us to systematically compare predictive performance when multiple binding models for the same TF are available. We introduce PyProBound, an easily extensible reimplementation of the ProBound biophysically interpretable machine learning framework. Configuring PyProBound to explicitly account for a confounding sequence-specific bias in DNA fragmentation rate yields improved TF binding models when training on ChIP-seq data. We also show how our likelihood function can be leveraged to perform de novo motif discovery on the raw allele-aware ChIP-seq counts.

Allele-specific binding (ASB) analyzer for annotation of allele-specific binding SNPs

BACKGROUND

Allele-specific binding (ASB) events occur when transcription factors (TFs) bind more favorably to one of the two parental alleles at heterozygous single nucleotide polymorphisms (SNPs). Evidence suggests that ASB events could reveal the impact of sequence variations on TF binding and may have implications for the risk of diseases.

RESULTS

Here we present ASB-analyzer, a software platform that enables the users to quickly and efficiently input raw sequencing data to generate individual reports containing the cytogenetic map of ASB SNPs and their associated phenotypes. This interactive tool thereby combines ASB SNP identification, biological annotation, motif analysis, phenotype associations and report summary in one pipeline. With this pipeline, we identified 3772 ASB SNPs from thirty GM12878 ChIP-seq datasets and demonstrated that the ASB SNPs were more likely to be enriched at important sites in TF-binding domains.

CONCLUSIONS

ASB-analyzer is a user-friendly tool that enables the detection, characterization and visualization of ASB SNPs. It is implemented in Python, R and bash shell and packaged in the Conda environment. It is available as an open-source tool on GitHub at https://github.com/Liying1996/ASBanalyzer .

Integration of risk variants from GWAS with SARS-CoV-2 RNA interactome prioritizes FUBP1 and RAB2A as risk genes for COVID-19

The role of host genetic factors in COVID-19 outcomes remains unclear despite various genome-wide association studies (GWAS). We annotate all significant variants and those variants in high LD (R2 > 0.8) from the COVID-19 host genetics initiative (HGI) and identify risk genes by recognizing genes intolerant nonsynonymous mutations in coding regions and genes associated with cis-expression quantitative trait loci (cis-eQTL) in non-coding regions. These genes are enriched in the immune response pathway and viral life cycle. It has been found that host RNA binding proteins (RBPs) participate in different phases of the SARS-CoV-2 life cycle. We collect 503 RBPs that interact with SARS-CoV-2 RNA concluded from in vitro studies. Combining risk genes from the HGI with RBPs, we identify two COVID-19 risk loci that regulate the expression levels of FUBP1 and RAB2A in the lung. Due to the risk allele, COVID-19 patients show downregulation of FUBP1 and upregulation of RAB2A. Using single-cell RNA sequencing data, we show that FUBP1 and RAB2A are expressed in SARS-CoV-2-infected upper respiratory tract epithelial cells. We further identify NC_000001.11:g.77984833C>A and NC_000008.11:g.60559280T>C as functional variants by surveying allele-specific transcription factor sites and cis-regulatory elements and performing motif analysis. To sum up, our research, which associates human genetics with expression levels of RBPs, identifies FUBP1 and RAB2A as two risk genes for COVID-19 and reveals the anti-viral role of FUBP1 and the pro-viral role of RAB2A in the infection of SARS-CoV-2.

A Bayesian method to infer copy number clones from single-cell RNA and ATAC sequencing

Single-cell RNA and ATAC sequencing technologies enable the examination of gene expression and chromatin accessibility in individual cells, providing insights into cellular phenotypes. In cancer research, it is important to consistently analyze these states within an evolutionary context on genetic clones. Here we present CONGAS+, a Bayesian model to map single-cell RNA and ATAC profiles onto the latent space of copy number clones. CONGAS+ clusters cells into tumour subclones with similar ploidy, rendering straightforward to compare their expression and chromatin profiles. The framework, implemented on GPU and tested on real and simulated data, scales to analyse seamlessly thousands of cells, demonstrating better performance than single-molecule models, and supporting new multi-omics assays. In prostate cancer, lymphoma and basal cell carcinoma, CONGAS+ successfully identifies complex subclonal architectures while providing a coherent mapping between ATAC and RNA, facilitating the study of genotype-phenotype maps and their connection to genomic instability.

Allele biased transcription factor binding across human brain regions gives mechanistic insight into eQTLs

Transcription Factors (TFs) influence gene expression by facilitating or disrupting the formation of transcription initiation machinery at particular genomic loci. Because genomic localization of TFs is in part driven by TF recognition of DNA sequence, variation in TF binding sites can disrupt TF-DNA associations and affect gene regulation. To identify variants that impact TF binding in human brain tissues, we quantified allele bias for 93 TFs analyzed with ChIP-seq experiments of multiple structural brain regions from two donors. Using graph genomes constructed from phased genomic sequence data, we compared ChIP-seq signal between alleles at heterozygous variants within each tissue sample from each donor. Comparison of results from different brain regions within donors and the same regions between donors provided measures of allele bias reproducibility. We identified thousands of DNA variants that show reproducible bias in ChIP-seq for at least one TF. We found that alleles that are rarer in the general population were more likely than common alleles to exhibit large biases, and more frequently led to reduced TF binding. Combining ChIP-seq with RNA-seq, we identified TF-allele interaction biases with RNA bias in a phased allele linked to 6,709 eQTL variants identified in GTEx data, 3,309 of which were found in neural contexts. Our results provide insights into the effects of both common and rare variation on gene regulation in the brain. These findings can facilitate mechanistic understanding of cis-regulatory variation associated with biological traits, including disease.

Transcription factor-binding k-mer analysis clarifies the cell type dependency of binding specificities and cis-regulatory SNPs in humans

BACKGROUND

Transcription factors (TFs) exhibit heterogeneous DNA-binding specificities in individual cells and whole organisms under natural conditions, and de novo motif discovery usually provides multiple motifs, even from a single chromatin immunoprecipitation-sequencing (ChIP-seq) sample. Despite the accumulation of ChIP-seq data and ChIP-seq-derived motifs, the diversity of DNA-binding specificities across different TFs and cell types remains largely unexplored.

RESULTS

Here, we applied MOCCS2, our k-mer-based motif discovery method, to a collection of human TF ChIP-seq samples across diverse TFs and cell types, and systematically computed profiles of TF-binding specificity scores for all k-mers. After quality control, we compiled a set of TF-binding specificity score profiles for 2,976 high-quality ChIP-seq samples, comprising 473 TFs and 398 cell types. Using these high-quality samples, we confirmed that the k-mer-based TF-binding specificity profiles reflected TF- or TF-family dependent DNA-binding specificities. We then compared the binding specificity scores of ChIP-seq samples with the same TFs but with different cell type classes and found that half of the analyzed TFs exhibited differences in DNA-binding specificities across cell type classes. Additionally, we devised a method to detect differentially bound k-mers between two ChIP-seq samples and detected k-mers exhibiting statistically significant differences in binding specificity scores. Moreover, we demonstrated that differences in the binding specificity scores between k-mers on the reference and alternative alleles could be used to predict the effect of variants on TF binding, as validated by in vitro and in vivo assay datasets. Finally, we demonstrated that binding specificity score differences can be used to interpret disease-associated non-coding single-nucleotide polymorphisms (SNPs) as TF-affecting SNPs and provide candidates responsible for TFs and cell types.

CONCLUSIONS

Our study provides a basis for investigating the regulation of gene expression in a TF-, TF family-, or cell-type-dependent manner. Furthermore, our differential analysis of binding-specificity scores highlights noncoding disease-associated variants in humans.

Allele specific binding of histone modifications and a transcription factor does not predict allele specific expression in correlated ChIP-seq peak-exon pairs

Allele specific expression (ASE) is widespread in many species including cows. Therefore, regulatory regions which control gene expression should show cis-regulatory variation which mirrors this differential expression within the animal. ChIP-seq peaks for histone modifications and transcription factors measure activity at functional regions and the height of some peaks have been shown to correlate across tissues with the expression of particular genes, suggesting these peaks are putative regulatory regions. In this study we identified ASE in the bovine genome in multiple tissues and investigated whether ChIP-seq peaks for four histone modifications and the transcription factor CTCF show allele specific binding (ASB) differences in the same tissues. We then investigate whether peak height and gene expression, which correlates across tissues, also correlates within the animal by investigating whether the direction of ASB in putative regulatory regions, mirrors that of the ASE in the genes they are putatively regulating. We found that ASE and ASB were widespread in the bovine genome but vary in extent between tissues. However, even when the height of a peak was positively correlated across tissues with expression of an exon, ASE of the exon and ASB of the peak were in the same direction only half the time. A likely explanation for this finding is that the correlations between peak height and exon expression do not indicate that the height of the peak causes the extent of exon expression, at least in some cases.

Analysis of regulatory network modules in hundreds of human stem cell lines reveals complex epigenetic and genetic factors contribute to pluripotency state differences between subpopulations

Stem cells exist in vitro in a spectrum of interconvertible pluripotent states. Analyzing hundreds of hiPSCs derived from different individuals, we show the proportions of these pluripotent states vary considerably across lines. We discovered 13 gene network modules (GNMs) and 13 regulatory network modules (RNMs), which were highly correlated with each other suggesting that the coordinated co-accessibility of regulatory elements in the RNMs likely underlied the coordinated expression of genes in the GNMs. Epigenetic analyses revealed that regulatory networks underlying self-renewal and pluripotency have a surprising level of complexity. Genetic analyses identified thousands of regulatory variants that overlapped predicted transcription factor binding sites and were associated with chromatin accessibility in the hiPSCs. We show that the master regulator of pluripotency, the NANOG-OCT4 Complex, and its associated network were significantly enriched for regulatory variants with large effects, suggesting that they may play a role in the varying cellular proportions of pluripotency states between hiPSCs. Our work captures the coordinated activity of tens of thousands of regulatory elements in hiPSCs and bins these elements into discrete functionally characterized regulatory networks, shows that regulatory elements in pluripotency networks harbor variants with large effects, and provides a rich resource for future pluripotent stem cell research.

rs71327024 Associated with COVID-19 Hospitalization Reduces CXCR6 Promoter Activity in Human CD4+ T Cells via Disruption of c-Myb Binding

Single-nucleotide polymorphism rs71327024 located in the human 3p21.31 locus has been associated with an elevated risk of hospitalization upon SARS-CoV-2 infection. The 3p21.31 locus contains several genes encoding chemokine receptors potentially relevant to severe COVID-19. In particular, CXCR6, which is prominently expressed in T lymphocytes, NK, and NKT cells, has been shown to be involved in the recruitment of immune cells to non-lymphoid organs in chronic inflammatory and respiratory diseases. In COVID-19, CXCR6 expression is reduced in lung resident memory T cells from patients with severe disease as compared to the control cohort with moderate symptoms. We demonstrate here that rs71327024 is located within an active enhancer that augments the activity of the CXCR6 promoter in human CD4+ T lymphocytes. The common rs71327024(G) variant makes a functional binding site for the c-Myb transcription factor, while the risk rs71327024(T) variant disrupts c-Myb binding and reduces the enhancer activity. Concordantly, c-Myb knockdown in PMA-treated Jurkat cells negates rs71327024's allele-specific effect on CXCR6 promoter activity. We conclude that a disrupted c-Myb binding site may decrease CXCR6 expression in T helper cells of individuals carrying the minor rs71327024(T) allele and thus may promote the progression of severe COVID-19 and other inflammatory pathologies.

Hidden modes of DNA binding by human nuclear receptors

Human nuclear receptors (NRs) are a superfamily of ligand-responsive transcription factors that have central roles in cellular function. Their malfunction is linked to numerous diseases, and the ability to modulate their activity with synthetic ligands has yielded 16% of all FDA-approved drugs. NRs regulate distinct gene networks, however they often function from genomic sites that lack known binding motifs. Here, to annotate genomic binding sites of known and unexamined NRs more accurately, we use high-throughput SELEX to comprehensively map DNA binding site preferences of all full-length human NRs, in complex with their ligands. Furthermore, to identify non-obvious binding sites buried in DNA-protein interactomes, we develop MinSeq Find, a search algorithm based on the MinTerm concept from electrical engineering and digital systems design. The resulting MinTerm sequence set (MinSeqs) reveal a constellation of binding sites that more effectively annotate NR-binding profiles in cells. MinSeqs also unmask binding sites created or disrupted by 52,106 single-nucleotide polymorphisms associated with human diseases. By implicating druggable NRs as hidden drivers of multiple human diseases, our results not only reveal new biological roles of NRs, but they also provide a resource for drug-repurposing and precision medicine.

Discovery of target genes and pathways at GWAS loci by pooled single-cell CRISPR screens

Most variants associated with complex traits and diseases identified by genome-wide association studies (GWAS) map to noncoding regions of the genome with unknown effects. Using ancestrally diverse, biobank-scale GWAS data, massively parallel CRISPR screens, and single-cell transcriptomic and proteomic sequencing, we discovered 124 cis-target genes of 91 noncoding blood trait GWAS loci. Using precise variant insertion through base editing, we connected specific variants with gene expression changes. We also identified trans-effect networks of noncoding loci when cis target genes encoded transcription factors or microRNAs. Networks were themselves enriched for GWAS variants and demonstrated polygenic contributions to complex traits. This platform enables massively parallel characterization of the target genes and mechanisms of human noncoding variants in both cis and trans.

Candidate SNP Markers Significantly Altering the Affinity of TATA-Binding Protein for the Promoters of Human Hub Genes for Atherogenesis, Atherosclerosis and Atheroprotection

Atherosclerosis is a systemic disease in which focal lesions in arteries promote the build-up of lipoproteins and cholesterol they are transporting. The development of atheroma (atherogenesis) narrows blood vessels, reduces the blood supply and leads to cardiovascular diseases. According to the World Health Organization (WHO), cardiovascular diseases are the leading cause of death, which has been especially boosted since the COVID-19 pandemic. There is a variety of contributors to atherosclerosis, including lifestyle factors and genetic predisposition. Antioxidant diets and recreational exercises act as atheroprotectors and can retard atherogenesis. The search for molecular markers of atherogenesis and atheroprotection for predictive, preventive and personalized medicine appears to be the most promising direction for the study of atherosclerosis. In this work, we have analyzed 1068 human genes associated with atherogenesis, atherosclerosis and atheroprotection. The hub genes regulating these processes have been found to be the most ancient. In silico analysis of all 5112 SNPs in their promoters has revealed 330 candidate SNP markers, which statistically significantly change the affinity of the TATA-binding protein (TBP) for these promoters. These molecular markers have made us confident that natural selection acts against underexpression of the hub genes for atherogenesis, atherosclerosis and atheroprotection. At the same time, upregulation of the one for atheroprotection promotes human health.

Semi-supervised learning improves regulatory sequence prediction with unlabeled sequences

MOTIVATION

Genome-wide association studies have systematically identified thousands of single nucleotide polymorphisms (SNPs) associated with complex genetic diseases. However, the majority of those SNPs were found in non-coding genomic regions, preventing the understanding of the underlying causal mechanism. Predicting molecular processes based on the DNA sequence represents a promising approach to understand the role of those non-coding SNPs. Over the past years, deep learning was successfully applied to regulatory sequence prediction using supervised learning. Supervised learning required DNA sequences associated with functional data for training, whose amount is strongly limited by the finite size of the human genome. Conversely, the amount of mammalian DNA sequences is exponentially increasing due to ongoing large sequencing projects, but without functional data in most cases.

RESULTS

To alleviate the limitations of supervised learning, we propose a paradigm shift with semi-supervised learning, which does not only exploit labeled sequences (e.g. human genome with ChIP-seq experiment), but also unlabeled sequences available in much larger amounts (e.g. from other species without ChIP-seq experiment, such as chimpanzee). Our approach is flexible and can be plugged into any neural architecture including shallow and deep networks, and shows strong predictive performance improvements compared to supervised learning in most cases (up to [Formula: see text]).

AVAILABILITY AND IMPLEMENTATION

https://forgemia.inra.fr/raphael.mourad/deepgnn .

Current sequence-based models capture gene expression determinants in promoters but mostly ignore distal enhancers

BACKGROUND

The largest sequence-based models of transcription control to date are obtained by predicting genome-wide gene regulatory assays across the human genome. This setting is fundamentally correlative, as those models are exposed during training solely to the sequence variation between human genes that arose through evolution, questioning the extent to which those models capture genuine causal signals.

RESULTS

Here we confront predictions of state-of-the-art models of transcription regulation against data from two large-scale observational studies and five deep perturbation assays. The most advanced of these sequence-based models, Enformer, by and large, captures causal determinants of human promoters. However, models fail to capture the causal effects of enhancers on expression, notably in medium to long distances and particularly for highly expressed promoters. More generally, the predicted impact of distal elements on gene expression predictions is small and the ability to correctly integrate long-range information is significantly more limited than the receptive fields of the models suggest. This is likely caused by the escalating class imbalance between actual and candidate regulatory elements as distance increases.

CONCLUSIONS

Our results suggest that sequence-based models have advanced to the point that in silico study of promoter regions and promoter variants can provide meaningful insights and we provide practical guidance on how to use them. Moreover, we foresee that it will require significantly more and particularly new kinds of data to train models accurately accounting for distal elements.

Altered and allele-specific open chromatin landscape reveals epigenetic and genetic regulators of innate immunity in COVID-19

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection causes severe COVID-19 in some patients and mild COVID-19 in others. Dysfunctional innate immune responses have been identified to contribute to COVID-19 severity, but the key regulators are still unknown. Here, we present an integrative single-cell multi-omics analysis of peripheral blood mononuclear cells from hospitalized and convalescent COVID-19 patients. In classical monocytes, we identified genes that were potentially regulated by differential chromatin accessibility. Then, sub-clustering and motif-enrichment analyses revealed disease condition-specific regulation by transcription factors and their targets, including an interaction between C/EBPs and a long-noncoding RNA LUCAT1, which we validated through loss-of-function experiments. Finally, we investigated genetic risk variants that exhibit allele-specific open chromatin (ASoC) in COVID-19 patients and identified a SNP rs6800484-C, which is associated with lower expression of CCR2 and may contribute to higher viral loads and higher risk of COVID-19 hospitalization. Altogether, our study highlights the diverse genetic and epigenetic regulators that contribute to COVID-19.

Genetic adaptation to pathogens and increased risk of inflammatory disorders in post-Neolithic Europe

Ancient genomics can directly detect human genetic adaptation to environmental cues. However, it remains unclear how pathogens have exerted selective pressures on human genome diversity across different epochs and affected present-day inflammatory disease risk. Here, we use an ancestry-aware approximate Bayesian computation framework to estimate the nature, strength, and time of onset of selection acting on 2,879 ancient and modern European genomes from the last 10,000 years. We found that the bulk of genetic adaptation occurred after the start of the Bronze Age, <4,500 years ago, and was enriched in genes relating to host-pathogen interactions. Furthermore, we detected directional selection acting on specific leukocytic lineages and experimentally demonstrated that the strongest negatively selected candidate variant in immunity genes, lipopolysaccharide-binding protein (LBP) D283G, is hypomorphic. Finally, our analyses suggest that the risk of inflammatory disorders has increased in post-Neolithic Europeans, possibly because of antagonistic pleiotropy following genetic adaptation to pathogens.

An atlas of transcribed human cardiac promoters and enhancers reveals an important role of regulatory elements in heart failure

A deeper knowledge of the dynamic transcriptional activity of promoters and enhancers is needed to improve mechanistic understanding of the pathogenesis of heart failure and heart diseases. In this study, we used cap analysis of gene expression (CAGE) to identify and quantify the activity of transcribed regulatory elements (TREs) in the four cardiac chambers of 21 healthy and ten failing adult human hearts. We identified 17,668 promoters and 14,920 enhancers associated with the expression of 14,519 genes. We showed how these regulatory elements are alternatively transcribed in different heart regions, in healthy versus failing hearts and in ischemic versus non-ischemic heart failure samples. Cardiac-disease-related single-nucleotide polymorphisms (SNPs) appeared to be enriched in TREs, potentially affecting the allele-specific transcription factor binding. To conclude, our open-source heart CAGE atlas will serve the cardiovascular community in improving the understanding of the role of the cardiac gene regulatory networks in cardiovascular disease and therapy.

Optimized high-throughput screening of non-coding variants identified from genome-wide association studies

The vast majority of disease-associated single nucleotide polymorphisms (SNP) identified from genome-wide association studies (GWAS) are localized in non-coding regions. A significant fraction of these variants impact transcription factors binding to enhancer elements and alter gene expression. To functionally interrogate the activity of such variants we developed snpSTARRseq, a high-throughput experimental method that can interrogate the functional impact of hundreds to thousands of non-coding variants on enhancer activity. snpSTARRseq dramatically improves signal-to-noise by utilizing a novel sequencing and bioinformatic approach that increases both insert size and the number of variants tested per loci. Using this strategy, we interrogated known prostate cancer (PCa) risk-associated loci and demonstrated that 35% of them harbor SNPs that significantly altered enhancer activity. Combining these results with chromosomal looping data we could identify interacting genes and provide a mechanism of action for 20 PCa GWAS risk regions. When benchmarked to orthogonal methods, snpSTARRseq showed a strong correlation with in vivo experimental allelic-imbalance studies whereas there was no correlation with predictive in silico approaches. Overall, snpSTARRseq provides an integrated experimental and computational framework to functionally test non-coding genetic variants.

Targeted profiling of human extrachromosomal DNA by CRISPR-CATCH

Extrachromosomal DNA (ecDNA) is a common mode of oncogene amplification but is challenging to analyze. Here, we adapt CRISPR-CATCH, in vitro CRISPR-Cas9 treatment and pulsed field gel electrophoresis of agarose-entrapped genomic DNA, previously developed for bacterial chromosome segments, to isolate megabase-sized human ecDNAs. We demonstrate strong enrichment of ecDNA molecules containing EGFR, FGFR2 and MYC from human cancer cells and NRAS ecDNA from human metastatic melanoma with acquired therapeutic resistance. Targeted enrichment of ecDNA versus chromosomal DNA enabled phasing of genetic variants, identified the presence of an EGFRvIII mutation exclusively on ecDNAs and supported an excision model of ecDNA genesis in a glioblastoma model. CRISPR-CATCH followed by nanopore sequencing enabled single-molecule ecDNA methylation profiling and revealed hypomethylation of the EGFR promoter on ecDNAs. We distinguished heterogeneous ecDNA species within the same sample by size and sequence with base-pair resolution and discovered functionally specialized ecDNAs that amplify select enhancers or oncogene-coding sequences.

Integrative Functional Genomics Identifies Systemic Lupus Erythematosus Causal Genetic Variant in the IRF5 Risk Locus

OBJECTIVE

IRF5 plays a crucial role in the development of lupus. Genome-wide association studies have identified several systemic lupus erythematosus (SLE) risk single-nucleotide polymorphisms (SNPs) enriched in the IRF5 locus. However, no comprehensive genome editing-based functional analysis exists to establish a direct link between these variants and altered IRF5 expression, particularly for enhancer variants. This study was undertaken to dissect the regulatory function and mechanisms of SLE IRF5 enhancer risk variants and to explore the utilization of clustered regularly interspaced short palindromic repeat interference (CRISPRi) to regulate the expression of disease risk gene to intervene in the disease.

METHODS

Epigenomic profiles and expression quantitative trait locus analysis were applied to prioritize putative functional variants in the IRF5 locus. CRISPR-mediated deletion, activation, and interference were performed to investigate the genetic function of rs4728142. Allele-specific chromatin immunoprecipitation-quantitative polymerase chain reaction and allele-specific formaldehyde-assisted isolation of regulatory element-quantitative polymerase chain reaction were used to decipher the mechanism of alleles differentially regulating IRF5 expression. The CRISPRi approach was used to evaluate the intervention effect in monocytes from SLE patients.

RESULTS

SLE risk SNP rs4728142 was located in an enhancer region, indicating a disease-related regulatory function, and risk allele rs4728142-A was closely associated with increased IRF5 expression. We demonstrated that an rs4728142-containing region could act as an enhancer to regulate the expression of IRF5. Moreover, rs4728142 affected the binding affinity of zinc finger and BTB domain-containing protein 3 (ZBTB3), a transcription factor involved in regulation. Furthermore, in monocytes from SLE patients, CRISPR-based interference with the regulation of this enhancer attenuated the production of disease-associated cytokines.

CONCLUSION

These results demonstrate that the rs4728142-A allele increases the SLE risk by affecting ZBTB3 binding, chromatin status, and regulating IRF5 expression, establishing a biologic link between genetic variation and lupus pathogenesis.

ANANASTRA: annotation and enrichment analysis of allele-specific transcription factor binding at SNPs

We present ANANASTRA, https://ananastra.autosome.org, a web server for the identification and annotation of regulatory single-nucleotide polymorphisms (SNPs) with allele-specific binding events. ANANASTRA accepts a list of dbSNP IDs or a VCF file and reports allele-specific binding (ASB) sites of particular transcription factors or in specific cell types, highlighting those with ASBs significantly enriched at SNPs in the query list. ANANASTRA is built on top of a systematic analysis of allelic imbalance in ChIP-Seq experiments and performs the ASB enrichment test against background sets of SNPs found in the same source experiments as ASB sites but not displaying significant allelic imbalance. We illustrate ANANASTRA usage with selected case studies and expect that ANANASTRA will help to conduct the follow-up of GWAS in terms of establishing functional hypotheses and designing experimental verification.

Epigenetic and transcriptional dysregulation in CD4+ T cells in patients with atopic dermatitis

Atopic dermatitis (AD) is one of the most common skin disorders among children. Disease etiology involves genetic and environmental factors, with 29 independent AD risk loci enriched for risk allele-dependent gene expression in the skin and CD4+ T cell compartments. We investigated the potential epigenetic mechanisms responsible for the genetic susceptibility of CD4+ T cells. To understand the differences in gene regulatory activity in peripheral blood T cells in AD, we measured chromatin accessibility (an assay based on transposase-accessible chromatin sequencing, ATAC-seq), nuclear factor kappa B subunit 1 (NFKB1) binding (chromatin immunoprecipitation with sequencing, ChIP-seq), and gene expression levels (RNA-seq) in stimulated CD4+ T cells from subjects with active moderate-to-severe AD, as well as in age-matched non-allergic controls. Open chromatin regions in stimulated CD4+ T cells were highly enriched for AD genetic risk variants, with almost half of the AD risk loci overlapping AD-dependent ATAC-seq peaks. AD-specific open chromatin regions were strongly enriched for NF-κB DNA-binding motifs. ChIP-seq identified hundreds of NFKB1-occupied genomic loci that were AD- or control-specific. As expected, the AD-specific ChIP-seq peaks were strongly enriched for NF-κB DNA-binding motifs. Surprisingly, control-specific NFKB1 ChIP-seq peaks were not enriched for NFKB1 motifs, but instead contained motifs for other classes of human transcription factors, suggesting a mechanism involving altered indirect NFKB1 binding. Using DNA sequencing data, we identified 63 instances of altered genotype-dependent chromatin accessibility at 36 AD risk variant loci (30% of AD risk loci) that might lead to genotype-dependent gene expression. Based on these findings, we propose that CD4+ T cells respond to stimulation in an AD-specific manner, resulting in disease- and genotype-dependent chromatin accessibility alterations involving NFKB1 binding.

Functional Characterization of Genetic Variant Effects on Expression

Thousands of common genetic variants in the human population have been associated with disease risk and phenotypic variation by genome-wide association studies (GWAS). However, the majority of GWAS variants fall into noncoding regions of the genome, complicating our understanding of their regulatory functions, and few molecular mechanisms of GWAS variant effects have been clearly elucidated. Here, we set out to review genetic variant effects, focusing on expression quantitative trait loci (eQTLs), including their utility in interpreting GWAS variant mechanisms. We discuss the interrelated challenges and opportunities for eQTL analysis, covering determining causal variants, elucidating molecular mechanisms of action, and understanding context variability. Addressing these questions can enable better functional characterization of disease-associated loci and provide insights into fundamental biological questions of the noncoding genetic regulatory code and its control of gene expression. Expected final online publication date for the Annual Review of Biomedical Data Science, Volume 5 is August 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Lupus enhancer risk variant causes dysregulation of IRF8 through cooperative lncRNA and DNA methylation machinery

Despite strong evidence that human genetic variants affect the expression of many key transcription factors involved in autoimmune diseases, establishing biological links between non-coding risk variants and the gene targets they regulate remains a considerable challenge. Here, we combine genetic, epigenomic, and CRISPR activation approaches to screen for functional variants that regulate IRF8 expression. We demonstrate that the locus containing rs2280381 is a cell-type-specific enhancer for IRF8 that spatially interacts with the IRF8 promoter. Further, rs2280381 mediates IRF8 expression through enhancer RNA AC092723.1, which recruits TET1 to the IRF8 promoter regulating IRF8 expression by affecting methylation levels. The alleles of rs2280381 modulate PU.1 binding and chromatin state to regulate AC092723.1 and IRF8 expression differentially. Our work illustrates an integrative strategy to define functional genetic variants that regulate the expression of critical genes in autoimmune diseases and decipher the mechanisms underlying the dysregulation of IRF8 expression mediated by lupus risk variants.

Functional analysis of the 1p34.3 risk locus implicates GNL2 in high-grade serous ovarian cancer

The high-grade serous ovarian cancer (HGSOC) risk locus at chromosome 1p34.3 resides within a frequently amplified genomic region signifying the presence of an oncogene. Here, we integrate in silico variant-to-function analysis with functional studies to characterize the oncogenic potential of candidate genes in the 1p34.3 locus. Fine mapping of genome-wide association statistics identified candidate causal SNPs local to H3K27ac-demarcated enhancer regions that exhibit allele-specific binding for CTCF in HGSOC and normal fallopian tube secretory epithelium cells (FTSECs). SNP risk associations colocalized with eQTL for six genes (DNALI1, GNL2, RSPO1, SNIP1, MEAF6, and LINC01137) that are more highly expressed in carriers of the risk allele, and three (DNALI1, GNL2, and RSPO1) were upregulated in HGSOC compared to normal ovarian surface epithelium cells and/or FTSECs. Increased expression of GNL2 and MEAF6 was associated with shorter survival in HGSOC with 1p34.3 amplifications. Despite its activation of β-catenin signaling, RSPO1 overexpression exerted no effects on proliferation or colony formation in our study of ovarian cancer and FTSECs. Instead, GNL2, MEAF6, and SNIP1 silencing impaired in vitro ovarian cancer cell growth. Additionally, GNL2 silencing diminished xenograft tumor formation, whereas overexpression stimulated proliferation and colony formation in FTSECs. GNL2 influences 60S ribosomal subunit maturation and global protein synthesis in ovarian cancer and FTSECs, providing a potential mechanism of how GNL2 upregulation might promote ovarian cancer development and mediate genetic susceptibility of HGSOC.