Kumardeep Chaudhary

A Comparative Benchmark of Biomedical Language Models for Concept Normalization from Real-World Text

This study benchmarks biomedical language models for concept normalization from real-world text, finding Qwen 2 Instruct improves accuracy when combined with SapBERT.

Whole-Genome Landscape of Breast Cancers from India shows Distinct Clinically Actionable Subtypes

This study presents whole-genome and transcriptome analyses of breast tumors from India, identifying novel mutations, copy number alterations, and subtypes with clinical implications.

DeepProg: an ensemble of deep-learning and machine-learning models for prognosis prediction using multi-omics data

DeepProg predicts patient survival subtypes using multi-omics data, identifying two optimal subtypes and improving risk-stratification across multiple cancers.

Ancestrally and Temporally Diverse Analysis of Penetrance of Clinical Variants in 72,434 Individuals

The paper quantifies disease risk associated with 37,772 clinical variants across 72,434 individuals, revealing heterogeneity in penetrance and limitations of current variant classification systems.

Exome-by-phenome-wide rare variant gene burden association with electronic health record phenotypes

This study identifies robust gene-phenotype associations by analyzing rare loss-of-function variants in exome-wide data linked to electronic health records.

Identification of Discriminative Gene-level and Protein-level Features Associated with Gain-of-Function and Loss-of-Function Mutations

The paper identifies gene- and protein-level features distinguishing gain-of-function and loss-of-function mutations using NLP and machine learning on a new database of pathogenic mutations.

Unsupervised Machine learning to subtype Sepsis-Associated Acute Kidney Injury

This study uses unsupervised machine learning on EHR data to identify two subphenotypes of sepsis-associated AKI with differing mortality rates, emphasizing the importance of intra-patient variability.

Computer-aided prediction of antigen presenting cell modulators for designing peptide-based vaccine adjuvants

The paper develops machine learning models to predict antigen presenting cell modulators (A-cell epitopes) and implements them in a web server for vaccine adjuvant design.

Multi-modal meta-analysis of 1494 hepatocellular carcinoma samples reveals vast impacts of consensus driver genes on phenotypes

This study identifies 10 consensus driver genes in HCC across six cohorts and shows their associations with mRNA transcriptomes, patient survival, and demographic factors.

A web bench for analysis and prediction of oncological status from proteomics data of urine samples

The paper develops proteomics-based prediction models to distinguish urinary tract oncological samples from healthy controls using urine data and creates a web tool for data mining.

Deep learning accurately predicts estrogen receptor status in breast cancer metabolomics data

This study compares deep learning and machine learning methods for classifying estrogen receptor status in breast cancer metabolomics data, finding deep learning achieves higher accuracy.

Deep Learning based multi-omics integration robustly predicts survival in liver cancer

This paper develops a deep learning model integrating multi-omics data to predict survival in liver cancer, identifying two patient subgroups with significant survival differences.