Sushant Kumar

The paper develops DeepND to analyze nucleosomal and linker DNA sequence patterns, identifying motifs linked to cancer mutations and methylation.

The paper presents an interpretable machine-learning framework to predict liquid-liquid phase separation in intrinsically disordered regions and assesses its relevance to disease variants.

The paper analyzes non-B DNA motifs in haplotype-resolved human genomes, revealing population-level variation and structural stability patterns in repetitive genomic regions.

The paper uses genomic and epigenomic data to identify mechanisms underlying the origin of structural variations in the human genome.

This paper introduces a machine learning tool for micronuclei segmentation and refines nuclei-micronuclei relationships using a two-phase labeling approach and a Mask-RCNN model.

The paper presents 64 haplotype-resolved human genomes using long-read sequencing and identifies structural variants, SV-eQTLs, and adaptive selection candidates.

SVFX assigns pathogenicity scores to structural variants in cancer and cardiovascular disease cohorts using machine learning.

This paper identifies cancer driver genes by analyzing mutational hotspots in 3D protein structures, incorporating protein dynamics to improve sensitivity.

This study evaluates the reliability of whole-exome sequencing in detecting intratumor genetic heterogeneity, finding that technical noise accounts for a significant proportion of apparent genetic differences.

The paper analyzes 2500 cancer genomes to assess the molecular functional impact of mutations, finding that some putative passenger variants may influence gene activity and contribute to cancer phenotypes.

The paper identifies and characterizes coding and non-coding driver mutations across 2,500 cancer genomes, revealing new candidates and questioning previous findings.