Anshul Bharat Kundaje

The paper introduces CPA-Perturb-seq, a multiplexed perturbation screen dataset, and uses it to identify polyA site regulators and their functional properties.

The paper reports 109 non-coding CRISPR screens analyzing 13.3Mb of the genome, identifies CRE-gene links, and provides guidelines and tools for screening non-coding regulatory elements.

This study uses single-cell chromatin accessibility analysis to identify developmental trajectories and non-coding mutations associated with congenital heart disease.

The paper uses single-cell multi-omics to study how pathogenic mtDNA mutations are subject to purifying selection in immune cells from Pearson syndrome patients.

The paper uses a single-cell multiome atlas of the human retina and deep learning to identify causal noncoding variants in eye diseases.

This paper introduces Yuzu, a method that accelerates in-silico saturation mutagenesis by using compressed sensing, reducing computation time for convolution-based models.

The paper introduces a genome browser track for visualizing dynamic DNA sequence features at single nucleotide resolution using machine learning-derived scores.

The paper uses single-cell ATAC-seq to analyze coronary artery cells from CAD patients, identifying regulatory mechanisms and risk variants associated with disease.

The paper identifies vascular site-specific epigenomic and transcriptomic features and links them to disease risk loci using single-cell data and GWAS integration.

The paper introduces APARENT2, a neural network model for predicting polyadenylation effects, and finds gain-of-function mutations linked to Autism Spectrum Disorder.

This study improves cross-species prediction of transcription factor binding by addressing domain shifts caused by species-specific repeats using domain adaptive neural networks.

fastISM accelerates in-silico saturation mutagenesis for convolutional neural networks by 10x, making it feasible for large-scale sequence analysis.

This study maps open chromatin, transcription, and 3D organization of the nucleomorph genome in Bigelowiella natans, revealing its highly accessible and transcribed state with strong nucleosome positioning and telomeric chromosome interactions.

This paper benchmarks deep learning interpretation methods on more realistic simulated regulatory genomic data, finding that some methods perform poorly on complex datasets.

This paper develops a method to generate neuronal cells from adult fibroblasts and uses it to study gene expression and epigenetic changes in 22q11.2 deletion syndrome.

The paper identifies dynamic cis-regulatory elements in epidermal differentiation using multi-omic data and deep learning, revealing combinatorial motif rules linked to gene expression and polygenic skin disease.

The paper maps gene-regulatory dynamics in the developing human cerebral cortex using single-cell data, identifying key transcription factors and disrupted regulatory elements in ASD.

This study examines the 3D genome organization of dinoflagellates, finding that transcription-induced supercoiling drives domain-scale structure without chromatin loops.

The paper introduces a method to profile off-target specificity of CRISPR enzymes using ssDNA mapping, applicable to both catalytically dead and active Cas9 variants.

This paper develops a statistical method to identify gene co-essentiality, creating a genome-wide functional module almanac that predicts functions for 102 uncharacterized genes and validates key findings.

The paper develops a CRISPR-based mutagenesis approach combined with machine learning to model cis-regulatory features of ADAR-mediated RNA editing, explaining over 70% of editing level variation.

MTSplice predicts the effects of genetic variants on tissue-specific splicing in 56 tissues and identifies enrichment of autism-associated de novo mutations in brain splicing.

The paper uses single-cell data to show a continuum of cell state and composition changes from polyps to colorectal cancer, identifying stem-like cells and regulatory markers.

The paper introduces BPNet, a deep learning model that predicts base-resolution ChIP-nexus binding profiles of TFs and identifies soft syntax rules for cooperative TF binding.

The paper develops HT-recruit, a high-throughput assay to identify transcriptional effectors in human proteins, finding repressor function correlates with evolutionary age and discovering new activator domains.

This paper evaluates strategies for enforcing reverse-complement equivariance in deep learning models for regulatory genomics, finding post-hoc conjoined models perform best.

The paper uses single-cell epigenomic data to identify functional SNPs associated with Alzheimer's and Parkinson's disease, revealing novel gene-disease associations and regulatory mechanisms.

The paper uses the ABC Model to map enhancer-gene regulation across 131 cell types and tissues, linking GWAS variants to disease genes, particularly in IBD.

This paper introduces a Fourier-transform-based attribution prior to improve deep learning model stability, interpretability, and performance in genomics tasks.

This paper introduces SMAC-seq, a method for long-range chromatin accessibility mapping in eukaryotes, and demonstrates its application in yeast.