Biology subjects

Kundaje, A.

Publications and source records attributed to Kundaje, A..

At least 37 records · Page 2Linked to original sources

bioRxiv2025

Dissecting regulatory syntax in human development with scalable multiomics and deep learning

Transcription factors (TFs) establish cell identity during development by binding regulatory DNA in a sequence-specific manner, often promoting local chromatin accessibility, and regulating gene expression. Mapping accessible chromatin offers critical insights into transcriptional control, but available datasets for human development are restricted to bulk tissue, single organs, or single modalities. Here, we present the Human Development Multiomic Atlas (HDMA), a single-cell atlas of chromatin accessibility and gene expression from 817,740 fetal cells across 12 organs, spanning 203 cell types and over 1 million candidate cis-regulatory elements, many of which exhibit organ-specific in vivo enhancer activity. Deep learning models trained to predict accessibility from local DNA sequence unravel a comprehensive lexicon of motifs which influence accessibility, including composite motifs exhibiting distinct syntactic constraints predicted to mediate TF cooperativity. We identify "hard" syntactic rules requiring precise motif spacing and orientation, "soft" rules allowing flexible motif arrangements, and ubiquitous motifs that inhibit accessibility. Model-based interpretation of genetic variants reveals that disruption of motifs with positive and negative effects is associated with concordant effects on gene expression. Our work delineates how motif syntax governs cell type-specific chromatin accessibility and provides a foundational resource for decoding cis-regulatory logic and interpreting genetic variation during human development.

genomics↗

bioRxiv2025

PISA: a versatile interpretation tool for visualizing cis-regulatory rules in genomic data

Sequence-to-function neural networks learn cis-regulatory sequence rules driving many types of genomic data. Interpreting these models to relate the sequence rules to underlying biological processes remains challenging, especially for complex genomic readouts such as MNase-seq, which maps nucleosome occupancy but is confounded by experimental bias. We introduce pairwise influence by sequence attribution (PISA), an interpretation tool that combinatorially decodes which bases contributed to the readout at a specific genomic coordinate. PISA visualizes the effects of transcription factor motifs, detects undiscovered motifs with complex contribution patterns, and reveals experimental biases. By learning the bias for MNase-seq, PISA enables unprecedented nucleosome prediction models, allowing the de novo discovery of nucleosome-positioning motifs and their longrange chromatin effects, as well as the design of sequences with altered nucleosome configurations. These results show that PISA is a versatile tool that expands our ability to train and interpret sequence-to-function neural networks on genomics data and understand the underlying cis-regulatory code.

genomics↗

bioRxiv2025

Single-cell multiome and enhancer connectome of human retinal pigment epithelium and choroid nominate pathogenic variants in age-related macular degeneration

Age-related macular degeneration (AMD) is a leading cause of vision loss worldwide. Genome-wide association studies (GWAS) of AMD have identified dozens of risk loci that may house disease targets. However, variants at these loci are largely noncoding, making it difficult to assess their function and whether they are causal. Here, we present a single-cell gene expression and chromatin accessibility atlas of human retinal pigment epithelium (RPE) and choroid to systematically analyze both coding and noncoding variants implicated in AMD. We employ HiChIP and Activity-by-Contact modeling to map enhancers in these tissues and predict cell and gene targets of risk variants. We further perform allele-specific self-transcribing active regulatory region sequencing (STARR-seq) to functionally test variant activity in RPE cells, including in the context of complement activation. Our work nominates new pathogenic variants and mechanisms in AMD and offers a rich and accessible resource for studying diseases of the RPE and choroid.

genomics↗

bioRxiv2025

Mapping the regulatory effects of common and rare non-coding variants across cellular and developmental contexts in the brain and heart

Whole genome sequencing has identified over a billion non-coding variants in humans, while GWAS has revealed the non-coding genome as a significant contributor to disease. However, prioritizing causal common and rare non-coding variants in human disease, and understanding how selective pressures have shaped the non-coding genome, remains a significant challenge. Here, we predicted the effects of 15 million variants with deep learning models trained on single-cell ATAC-seq across 132 cellular contexts in adult and fetal brain and heart, producing nearly two billion context-specific predictions. Using these predictions, we distinguish candidate causal variants underlying human traits and diseases and their context-specific effects. While common variant effects are more cell-type-specific, rare variants exert more cell-type-shared regulatory effects, with selective pressures particularly targeting variants affecting fetal brain neurons. To prioritize de novo mutations with extreme regulatory effects, we developed FLARE, a context-specific functional genomic model of constraint. FLARE outperformed other methods in prioritizing case mutations from autism-affected families near syndromic autism-associated genes; for example, identifying mutation outliers near CNTNAP2 that would be missed by alternative approaches. Overall, our findings demonstrate the potential of integrating single-cell maps with population genetics and deep learning-based variant effect prediction to elucidate mechanisms of development and disease-ultimately, supporting the notion that genetic contributions to neurodevelopmental disorders are predominantly rare.

genomics↗

bioRxiv2024

An Expanded Registry of Candidate cis-Regulatory Elements for Studying Transcriptional Regulation

Mammalian genomes contain millions of regulatory elements that control the complex patterns of gene expression. Previously, The ENCODE consortium mapped biochemical signals across many cell types and tissues and integrated these data to develop a Registry of 0.9 million human and 300 thousand mouse candidate cis-Regulatory Elements (cCREs) annotated with potential functions1. We have expanded the Registry to include 2.35 million human and 927 thousand mouse cCREs, leveraging new ENCODE datasets and enhanced computational methods. This expanded Registry covers hundreds of unique cell and tissue types, providing a comprehensive understanding of gene regulation. Functional characterization data from assays like STARR-seq, MPRA, CRISPR perturbation, and transgenic mouse assays now cover over 90% of human cCREs, revealing complex regulatory functions. We identified thousands of novel silencer cCREs and demonstrated their dual enhancer/silencer roles in different cellular contexts. Integrating the Registry with other ENCODE annotations facilitates genetic variation interpretation and trait-associated gene identification, exemplified by discovering KLF1 as a novel causal gene for red blood cell traits. This expanded Registry is a valuable resource for studying the regulatory genome and its impact on health and disease.

genomics↗

bioRxiv2024

ChromBPNet: bias factorized, base-resolution deep learning models of chromatin accessibility reveal cis-regulatory sequence syntax, transcription factor footprints and regulatory variants

Despite extensive mapping of cis-regulatory elements (cREs) across cellular contexts with chromatin accessibility assays, the sequence syntax and genetic variants that regulate transcription factor (TF) binding and chromatin accessibility at context-specific cREs remain elusive. We introduce ChromBPNet, a deep learning DNA sequence model of base-resolution accessibility profiles that detects, learns and deconvolves assay-specific enzyme biases from regulatory sequence determinants of accessibility, enabling robust discovery of compact TF motif lexicons, cooperative motif syntax and precision footprints across assays and sequencing depths. Extensive benchmarks show that ChromBPNet, despite its lightweight design, is competitive with much larger contemporary models at predicting variant effects on chromatin accessibility, pioneer TF binding and reporter activity across assays, cell contexts and ancestry, while providing interpretation of disrupted regulatory syntax. ChromBPNet also helps prioritize and interpret regulatory variants that influence complex traits and rare diseases, thereby providing a powerful lens to decode regulatory DNA and genetic variation.

genetics↗

bioRxiv2024

Comparative atlas of genome-wide chromatin-associated protein co-occupancy

Accurate transcriptional regulation and chromatin dynamics requires the coordination and activity of chromatin-associated proteins (CAPs) at distinct loci. While the combinatorial activity of a select set of CAPs has been previously examined, these studies are limited by the underrepresentation of proteins and cell types explored, making it difficult to identify the global associations as well as the conservation of these associations across different cell types. Here, we performed 270 CAP chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-Seq) experiments in both K562 and HepG2 cancer cell lines and explored the relationship between cell identity and CAP co-association using three distinct approaches. We employed a machine learning algorithm to organize the genome-wide binding profiles into 56 and 70 interpretable co-association modules for HepG2 and K562 cell lines, respectively. We found CAP co-association modules are mostly cell type-specific, however those present in both cell lines are largely comprised of TFs from a single TF family and anchor to unique loci via lineage-specific factors. While enhancer-associated co-binding modules were largely composed of cell type-specific CAPs, we found regulatory activity at promoter-enhancer module contacts to be enriched for chromatin remodeling proteins. Additionally, we used colocalization information derived from co-association models in conjunction with neural network models of transcription factor (TF) activity to identify high-confidence candidate TF cooperative pairs. Finally, through comparing CAP enrichment in high occupancy target (HOT) regions in K562 and HepG2 cell lines, we found cell type-specific HOT sites, but not common HOT sites, are selectively enriched at high copy number loci. Overall, this study uncovers principles of sequence-level and large-scale CAP genomic organization and demonstrates how this contributes to cell type-specific regulatory mechanisms and cellular functions.

genomics↗

bioRxiv2024

Directionality of Transcriptional Regulatory Elements

Divergent transcription is a critical marker of active transcriptional regulatory elements (TREs), including enhancers and promoters, in mammals. However, distal elements with unidirectional transcriptional patterns are often overlooked, leaving their identity and function poorly understood. Here, we performed a systematic comparison between divergent and unidirectional elements, revealing their distinct architectural and functional features. Our analysis also shows that unidirectional elements have younger sequence ages and are under weaker evolutionary constraints than divergent elements, indicating that they may represent a unique category of genomic regulatory function with more recent origins. Notably, we observed that some transcription factors, including CTCF, AP1, SP, and NFY, exhibit dual roles in modulating the directionality of TREs, either activating or repressing nascent transcription in a position-dependent manner. Overall, the elucidation of directionality enhances our understanding of the diverse architectural models, functional features, evolutionary dynamics, and regulatory logic of TREs.

molecular biology↗

bioRxiv2024

Mapping enhancer-gene regulatory interactions from single-cell data

Mapping enhancers and their target genes in specific cell types is crucial for understanding gene regulation and human disease genetics. However, accurately predicting enhancer-gene regulatory interactions from single-cell datasets has been challenging. Here, we introduce a new family of classification models, scE2G, to predict enhancer-gene regulation. These models use features from single-cell ATAC-seq or multiomic RNA and ATAC-seq data and are trained on a CRISPR perturbation dataset including >10,000 evaluated element-gene pairs. We benchmark scE2G models against CRISPR perturbations, fine-mapped eQTLs, and GWAS variant-gene associations and demonstrate state-of-the-art performance at prediction tasks across multiple cell types and categories of perturbations. We apply scE2G to build maps of enhancer-gene regulatory interactions in heterogeneous tissues and interpret noncoding variants associated with complex traits, nominating regulatory interactions linking INPP4B and IL15 to lymphocyte counts. The scE2G models will enable accurate mapping of enhancer-gene regulatory interactions across thousands of diverse human cell types.

genetics↗

bioRxiv2024

A consensus variant-to-function score to functionally prioritize variants for disease

Identifying and functionally characterizing causal disease variants in genome-wide association studies remains a pressing challenge. Here, we construct a consensus variant-to-function (cV2F) score that assigns a single value to each common single-nucleotide variant in the genome, and helps to predict and characterize causal disease variants. The cV2F score leverages features reflecting variant-level experimentally and computationally predicted function (e.g. allelic imbalance and sequence-based deep learning models) and element-level function (e.g. predicted enhancers), and learns optimal combinations of features by training a gradient boosting model on GWAS fine-mapping results. The cV2F-annotated variants attained an AUPRC of 0.822 at identifying held-out fine-mapped variants. Variants with high cV2F scores are highly enriched for heritability (14.2x, s.e. 0.5) across 66 diseases/traits, are uniquely informative for disease heritability, and are highly predictive of variants implicated by reporter assays; cV2F substantially outperforms previous variant-to-function scores using all of these metrics. GWAS fine-mapping of 110 diseases/traits informed by cV2F identified 14.3% more confidently fine-mapped (PIP > 0.95) variants than non-functionally informed fine-mapping. We further constructed tissue/cell line-specific cV2F scores that prioritize variants based on regulatory potential in specific tissues/cell lines, attaining high heritability enrichment for tissue-related diseases/traits (15.6x, s.e. 2.3) while providing independent information (average correlation of 0.27 with the primary cV2F score). We highlight examples of GWAS loci for which cV2F pinpoints causal variants with high confidence and elucidates their functional role.

genetics↗

bioRxiv2024

GENCODE: massively expanding the lncRNA catalog through capture long-read RNA sequencing

Accurate and complete gene annotations are indispensable for understanding how genome sequences encode biological functions. For more than twenty years, the GENCODE consortium has developed reference annotations for the human and mouse genomes, becoming a foundation for biomedical and genomics communities worldwide. Nevertheless, collections of important yet poorly-understood gene classes like long non-coding RNAs (lncRNAs) remain incomplete and scattered across multiple, uncoordinated catalogs. To address this, GENCODE has undertaken the most comprehensive lncRNA annotation effort to date. This is founded on the manually supervised computational annotation of full-length targeted long-read sequencing, on matched embryonic and adult tissues, of orthologous regions in human and mouse. Altogether 17,931 human genes (140,268 transcripts) and 22,784 mouse genes (136,169 transcripts) have been added to the GENCODE catalog representing a 2-fold and 6-fold growth in transcripts, respectively - the greatest increase in the number of annotated human genes since the sequencing of the human genome. Our targeted design assigned human-mouse orthologs at a rate beyond previous studies, tripling the number of human disease-associated lncRNAs that have mouse orthologs. Novel lncRNA genes consistently exhibit biological signals of functionality, and they greatly enhance the functional interpretability of the human genome. While poorly expressed in bulk RNA-Seq samples, many of them are highly expressed in specific cell populations, maybe even contributing to cell-type determination. The expanded GENCODE lncRNA annotations mark a critical step toward deciphering the human and mouse genomes.

genomics↗

bioRxiv2024

Sequence-based TCR-Peptide Representations Using Cross-Epitope Contrastive Fine-tuning of Protein Language Models

Understanding T-Cell receptor (TCR) and epitope interactions is critical for advancing our knowledge of the human immune system. Traditional approaches that use sequence similarity or structure data often struggle to scale and generalize across diverse TCR/epitope interactions. To address these limitations, we introduce ImmuneCLIP, a contrastive fine-tuning method that leverages pre-trained protein language models to align TCR and epitope embeddings in a shared latent space. ImmuneCLIP is evaluated on epitope ranking and binding prediction tasks, where it consistently outperforms sequence-similarity based methods and existing deep learning models. Furthermore, ImmuneCLIP shows strong generalization capabilities even with limited training data, highlighting its potential for studying diverse immune interactions and uncovering patterns that improve our understanding of human immune recognition systems.

immunology↗

bioRxiv2024

Prediction and design of transcriptional repressor domains with large-scale mutational scans and deep learning

Regulatory proteins have evolved diverse repressor domains (RDs) to enable precise context-specific repression of transcription. However, our understanding of how sequence variation impacts the functional activity of RDs is limited. To address this gap, we generated a high-throughput mutational scanning dataset measuring the repressor activity of 115,000 variant sequences spanning more than 50 RDs in human cells. We identified thousands of clinical variants with loss or gain of repressor function, including TWIST1 HLH variants associated with Saethre-Chotzen syndrome and MECP2 domain variants associated with Rett syndrome. We also leveraged these data to annotate short linear interacting motifs (SLiMs) that are critical for repression in disordered RDs. Then, we designed a deep learning model called TENet (Transcriptional Effector Network) that integrates sequence, structure and biochemical representations of sequence variants to accurately predict repressor activity. We systematically tested generalization within and across domains with varying homology using the mutational scanning dataset. Finally, we employed TENet within a directed evolution sequence editing framework to tune the activity of both structured and disordered RDs and experimentally test thousands of designs. Our work highlights critical considerations for future dataset design and model training strategies to improve functional variant prioritization and precision design of synthetic regulatory proteins.

genomics↗

bioRxiv2024

Prediction and functional interpretation of inter-chromosomal genome architecture from DNA sequence with TwinC

Three-dimensional nuclear DNA architecture comprises well-studied intra-chromosomal (cis) folding and less characterized inter-chromosomal (trans) interfaces. Current predictive models of 3D genome folding can effectively infer pairwise cis-chromatin interactions from the primary DNA sequence but generally ignore trans contacts. There is an unmet need for robust models of trans-genome organization that provide insights into their underlying principles and functional relevance. We present TwinC, an interpretable convolutional neural network model that reliably predicts trans contacts measurable through proximity ligation-dependent (in situ and intact Hi-C) and independent (DNA SPRITE) genome-wide chromatin conformation assays.. TwinC uses a paired sequence design from replicate Hi-C experiments to learn single base pair relevance in trans interactions across two stretches of DNA. The method achieves high predictive accuracy (AUROC=0.80) on a cross-chromosomal test set from in situ and intact Hi-C experiments in heart tissue. Furthermore, we train TwinC using in situ Hi-C data from the widely used GM12878 cell line and validate its performance with orthogonal DNA SPRITE assay in the same cell type. Mechanistically, the neural network learns the importance of compartments, chromatin accessibility, clustered transcription factor binding and G-quadruplexes in forming trans contacts. In summary, TwinC models and interprets trans genome architecture, shedding light on this poorly understood aspect of gene regulation.

bioinformatics↗

bioRxiv2024

A cell and transcriptome atlas of the human arterial vasculature

Contiguous arterial segments show different propensities for different vascular pathologies, yet mechanisms explaining these fundamental differences remain unknown. We sought to build a transcriptomic, cellular, and spatial atlas of human arterial cells across multiple different arterial segments to understand these underlying differences. Analysis of multiple isogenic arterial segments from healthy donors reveals a significant stereotyped pattern of cell type-specific segmental heterogeneity in healthy arteries. Combining single cell analysis with spatial transcriptomic data reveals cellular heterogeneity not captured by commonly used cell-type marker genes. Determinants of arterial transcriptomic identities are predominantly encoded in fibroblasts and smooth muscle cells (SMC), and their differentially expressed genes are particularly enriched for different vascular disease-associated genetic risk- loci and risk-genes. Adventitial fibroblast-specific heterogeneity in gene expression coincides with a disproportionally large number of vascular disease genetic signals, suggesting a previously unrecognized role for this cell type in disease risk. Adult arterial cells from different segments cluster not by anatomical proximity, but by embryonic origin. Global regulon analysis of disease related segment-specific gene expression program in fibroblast and SMC enriches for binding sites of transcription factors that are developmental master regulators whose expression persists into adulthood, suggesting an important functional role of the same developmental master regulators in adult gene expression and disease. Lastly, non-coding transcriptomes across arterial cells contain extensive variation in lncRNAs expressed in cell type- and segment-specific patterns, rivaling heterogeneity in protein coding transcriptomes. Differentially expressed LncRNA demonstrate enrichment for non-coding genetic signals for vascular diseases, suggesting a potential global role of segmental specific LncRNAs in regulating inherited human vascular disease risk.

cell biology↗

bioRxiv2024

Mutagenesis Sensitivity Mapping of HumanEnhancers In Vivo

Distant-acting enhancers are central to human development. However, our limited understanding of their functional sequence features prevents the interpretation of enhancer mutations in disease. Here, we determined the functional sensitivity to mutagenesis of human developmental enhancers in vivo. Focusing on seven enhancers active in the developing brain, heart, limb and face, we created over 1700 transgenic mice for over 260 mutagenized enhancer alleles. Systematic mutation of 12-basepair blocks collectively altered each sequence feature in each enhancer at least once. We show that 69% of all blocks are required for normal in vivo activity, with mutations more commonly resulting in loss (60%) than in gain (9%) of function. Using predictive modeling, we annotated critical nucleotides at base-pair resolution. The vast majority of motifs predicted by these machine learning models (88%) coincided with changes to in vivo function, and the models showed considerable sensitivity, identifying 59% of all functional blocks. Taken together, our results reveal that human enhancers contain a high density of sequence features required for their normal in vivo function and provide a rich resource for further exploration of human enhancer logic.

genomics↗

bioRxiv2024

Flexible use of conserved motif vocabularies constrains genome access in cell type evolution

Cell types evolve into a hierarchy with related types grouped into families. How cell type diversification is constrained by the stable separation between families over vast evolutionary times remains unknown. Here, integrating single-nucleus multiomic sequencing and deep learning, we show that hundreds of sequence features (motifs) divide into distinct sets associated with accessible genomes of specific cell type families. This division is conserved across highly divergent, early-branching animals including flatworms and cnidarians. While specific interactions between motifs delineate cell type relationships within families, surprisingly, these interactions are not conserved between species. Consistently, while deep learning models trained on one species can predict accessibility of other species sequences, their predictions frequently rely on distinct, but synonymous, motif combinations. We propose that long-term stability of cell type families is maintained through genome access specified by conserved motif sets, or vocabularies, whereas cell types diversify through flexible use of motifs within each set.

evolutionary biology↗

bioRxiv2024

An updated compendium and reevaluation of the evidence for nuclear transcription factor occupancy over the mitochondrial genome

In most eukaryotes, mitochondrial organelles contain their own genome, usually circular, which is the remnant of the genome of the ancestral bacterial endosymbiont that gave rise to modern mitochondria. Mitochondrial genomes are dramatically reduced in their gene content due to the process of endosymbiotic gene transfer to the nucleus; as a result most mitochondrial proteins are encoded in the nucleus and imported into mitochondria. This includes the components of the dedicated mitochondrial transcription and replication systems and regulatory factors, which are entirely distinct from the information processing systems in the nucleus. However, since the 1990s several nuclear transcription factors have been reported to act in mitochondria, and previously we identified 8 human and 3 mouse transcription factors (TFs) with strong localized enrichment over the mitochondrial genome using ChIP-seq (Chromatin Immunoprecipitation) datasets from the second phase of the ENCODE (Encyclopedia of DNA Elements) Project Consortium. Here, we analyze the greatly expanded in the intervening decade ENCODE compendium of TF ChIP-seq datasets (a total of 6,153 ChIP experiments for 942 proteins, of which 763 are sequence-specific TFs) combined with interpretative deep learning models of TF occupancy to create a comprehensive compendium of nuclear TFs that show evidence of association with the mitochondrial genome. We find some evidence for chrM occupancy for 50 nuclear TFs and two other proteins, with bZIP TFs emerging as most likely to be playing a role in mitochondria. However, we also observe that in cases where the same TF has been assayed with multiple antibodies and ChIP protocols, evidence for its chrM occupancy is not always reproducible. In the light of these findings, we discuss the evidential criteria for establishing chrM occupancy and reevaluate the overall compendium of putative mitochondrial-acting nuclear TFs.

genomics↗