Murad Chowdhury

GenoSiS dynamically finds patient-matched cohorts across populations using genotype similarity search to improve clinical analyses.

STIX annotates structural variants using long-read data, achieving high accuracy and identifying common SVs in population-scale datasets.

This study identifies noncoding regulatory regions and their mutations in insulinomas, linking them to altered β-cell function and tumor development.

Samplot creates visualizations for structural variant validation and includes a machine learning tool to reduce false positives across multiple sequencing technologies.

This paper develops Samplot-ML, a CNN that reduces false positives in structural variant detection by 47% while retaining 97% of true positives using visualization-based training.

The paper uses ATAC-seq signal processing and recurrent neural networks to classify transcriptional activity from chromatin accessibility data.