Boris Oskotsky

GENOAR enables scalable, AI-compatible single-cell transcriptomic analysis of over 54 million cells from humans and mice, outperforming human experts in data retrieval and enabling new biological insights.

CDRPipe integrates single-cell and bulk transcriptomic perturbation data to improve drug repurposing predictions by combining complementary resources.

The paper investigates how age, sex, and APOE genotype affect hippocampal cell-type-specific gene expression and communication in mice, revealing sex-divergent effects of APOE4.

The paper introduces anticlustering to balance batches and minimize batch effects in high-throughput sequencing, showing improved balance compared to existing methods.

A novel method defines ethnicity-specific reference intervals for lab tests using EHR data, showing better prognostic value than existing intervals for serum creatinine and hemoglobin A1C.