Ruchir Rastogi

I am a postdoctoral scientist at Stanford University in Anshul Kundaje's lab, supported by the Warren Alpert Computational Biology & Artificial Intelligence fellowship. I previously completed my Ph.D. in Computer Science at UC Berkeley, coadvised by Nir Yosef and Nilah Ioannidis.

My research focuses on developing interpretable machine learning methods for genomics—both to uncover biological principles and to make clinically useful predictions. I am currently working in two areas. First, I am developing sequence-based models that decode long-range interactions in the regulatory genome and predict individual disease risk from personal genetic variants. Second, I am building computational methods to predict how cells in the immune system respond to various perturbations.

Email: rrastogi [at] stanford [dot] edu

Publications

Preprints

Fine-tuning sequence-to-expression models on personal genome and transcriptome data

Journal Papers

Critical assessment of missense variant effect predictors on disease-relevant variant data
Personal transcriptome variation is poorly explained by current genomic deep learning models
Cross-protein transfer learning substantially improves zero-shot prediction of disease variant effects
X-CAP improves pathogenicity prediction of stopgain variants
Modulation of macrophage functional polarity towards anti-inflammatory phenotype with plasmid DNA delivery in CD44 targeting hyaluronic acid nanoparticles

Conference Proceedings

Characterizing uncertainty in predictions of genomic sequence-to-activity models