Sudhir Kumar

Phylogenomic subsampling and upsampling for efficient evolutionary analyses of big data

This paper introduces a scalable phylogenomic framework that reduces computational costs while maintaining accuracy in evolutionary analyses.

De novo transformer modeling improves recovery of genetic cell types from sparse single-cell RNA sequencing

This study develops a sample-specific transformer model to improve genotype imputation from sparse single-cell RNA sequencing data, enhancing the recovery of genetic cell types.

Species phylogenetics using intraprotein residue dependence inferred with protein foundation models

This study compares RAM-NJ and SAM-NJ phylogenies using protein foundation models to reconstruct species relationships without conventional substitution models or multiple sequence alignments.

Integrating Phylogenies with Chronology to Assemble the Tree of Life

This paper introduces Chrono-STA, a method to integrate molecular phylogenies with chronological data to build a supertree, showing improved accuracy over existing methods for data with limited species overlap.

TRUHiC: A TRansformer-embedded U-2 Net to enhance Hi-C data for 3D chromatin structure characterization

TRUHiC enhances low-resolution Hi-C data using deep generative modeling to improve 3D chromatin structure characterization across cell lines and species.

Entrenchment and contingency in neutral protein evolution with epistasis

The paper shows that protein evolution with epistasis and purifying selection does not lead to entrenchment, and that rate variation is due to epistasis, not historical contingency.

Evolutionary sparse learning with paired species contrast reveals the shared genetic basis of convergent traits

This study uses evolutionary sparse learning to identify shared genetic bases for convergent traits in different species, finding common genes involved in C4 photosynthesis and echolocation.

MEGA12: Molecular Evolutionary Genetic Analysis version 12 for adaptive and green computing

MEGA12 adds filtered substitution-model selection, matching full-search choices in all 240 simulated nucleotide alignments, and adaptive bootstrapping with an 81% average speed-up, accelerating maximum-likelihood phylogenetics.

Discovering Fragile Clades And Causal Sequences In Phylogenomics By Evolutionary Sparse Learning

The paper introduces metrics to identify fragile clades and causal sequences in phylogenomics using evolutionary sparse learning, validated across fungi, plant, and animal datasets.

Split-Transformer Impute (STI): A Transformer Framework for Genotype Imputation

STI is a transformer-based model for genome-scale genotype imputation that achieves high accuracy and handles multi-allelic variants effectively.

TopHap: Rapid inference of key phylogenetic structures from common haplotypes in large genome collections with limited diversity

TopHap infers phylogenetic structures from common haplotypes in large genome collections, improving accuracy and speed for SARS-CoV-2 variant phylogeny.

TreeMap: A Structured Approach to Fine Mapping of eQTL Variants

TreeMap identifies causal eQTL variants by accounting for multisite effects and linkage, showing improved accuracy and enabling genome-wide analysis of long-range eQTL.

Using a GTR+Γ substitution model for dating sequence divergence when stationarity and time-reversibility assumptions are violated

This study evaluates the impact of violating time-reversibility and stationarity assumptions in molecular dating using simulated data and finds that biases are small and reducible with multiple calibrations.

PathFinder: Bayesian inference of clone migration histories in cancer

PathFinder infers cancer clone migration histories using Bayesian methods, showing improved accuracy with more clones and genetic variants.

Contextual Classifications of Cancer Driver Genes

This study introduces a computational method to classify cancer driver genes as oncogenes or tumor suppressors based on mutation patterns across 33 cancer types, identifying novel genes and revealing context-specific functions.