Featured
Deep learning framework detecting Intracranial Hemorrhage from brain CT scans using segmentation-based ResNet architecture. Engineered a PII-blind REST API with split-key encryption for clinical deployment.
Impact: Produced 98% accuracy in radiologist-based validation and reduced per-case processing time by 60% via clinical-grade AI inference pipeline
PyTorch
ResNet
Medical Imaging
REST API
Encryption
Computational colocalization pipeline integrating GWAS and bulk RNA-seq eQTL data to identify candidate causal genes in Diversity Outbred mice. Implemented conditional analysis with bootstrapping to validate concordance between signals.
Impact: Identified a novel lncRNA and 2 protein-coding genes providing mechanistic interpretation of BMD-GWAS loci — selected as Reviewer’s Choice Abstract at ASHG 2025
R
coloc
bnlearn
GWAS
eQTL
HPC/SLURM
Nextflow-based pipeline integrating long-read RNA sequencing with mass spectrometry to discover novel transcript isoforms and protein products absent from GENCODE annotations.
Impact: Identified novel isoforms absent from GENCODE by integrating 2 modalities (RNA-seq + targeted mass spectrometry) — published in JASMS 2024
Nextflow
Long-Read Seq
Mass Spectrometry
Proteomics
Pseudotime trajectory analysis on single-cell RNA-seq data characterizing gene expression dynamics along osteoblast and osteoclast differentiation — mapping the molecular choreography of bone formation.
Impact: Identified 5 novel genes likely important for osteoblast and osteoclast differentiation from BMSC scRNA-seq data
Slingshot
scRNA-seq
WGCNA
Bayesian Networks
High-throughput processing pipeline for Low-Pass Whole Genome Sequencing achieving 15% reduction in genotyping costs while maintaining fidelity for downstream haplotype calling and eQTL analysis.
Impact: Cut mouse genotyping costs 15% while preserving 98% accuracy in genotype calls for downstream haplotype calling and eQTL analysis
Python
Bash
SLURM
WGS
Imputation
Architected a standardized data governance framework enforcing FAIR principles across all research data assets — transforming an unstructured research environment into a reproducible, audit-ready ecosystem.
Impact: Standardized FAIR data-governance for shared NGS infrastructure across 5+ concurrent projects via automated metadata, provenance tracking, and CI/CD
FAIR Principles
Data Governance
ETL
Metadata Standards