Jun 2023 – Aug 2023 · 2 mos · South San Francisco, California, United States · On-site

• Investigated guide RNA (gRNA) scoring methods for different CRISPR technologies. Developed and integrated an R Package for gRNA scoring into the crisprVerse ecosystem.
• Improved deep learning models (TensorFlow, Python) using transfer learning for designing gRNAs for CRISPRa. Our method led to a 3.5x improvement in predicting the gRNA activity scores on the internal datasets compared to the existing state-of-the-art models.

Sep 2019 – May 2025 · 5 yrs 8 mos · College Park

• Developed algorithmic tools, statistical methods and pipelines for processing, quantifying and analyzing multi-omic data such as bulk RNA-Seq, single-cell ATAC-Seq, RNA-Seq and long-read RNA-Seq data.
• Conceptualized and developed a novel data-driven framework for incorporating the uncertainty associated with RNA-Seq abundance estimates by organizing transcripts in a tree (TreeTerminus (Rust), beaveR (R package)). The transcript groups from the tree provide an alternative to genes/ transcripts features while controlling for uncertainty.
• Designed an uncertainty-aware tree-based differential testing method for RNA-Seq with the R package mehenDi. mehenDi increased power by 4% when compared to transcript-level differential analysis while controlling for false positives, and identified novel features.
• Led the development of a processing and mapping method for single-cell ATAC-Seq data and integrated it into piscem(Rust and C++). Our method uses 2.5x less memory and is 1.5x faster than the state-of-the-art while retaining a similar accuracy.
• Characterized cell types and studied alternative splicing using transcriptomic segments (Yanagi) for the single-cell data generated using the Smart-seq2 protocol and developed an R package for working with the Yanagi segment counts.

Aug 2019 – Apr 2025 · 5 yrs 8 mos · College Park

• Introduction to Data Science, Object Oriented Programming

Jan 2018 – Apr 2018 · 3 mos

• Introduction to Systems Biology

Aug 2017 – Apr 2019 · 1 yr 8 mos

• Developed a machine learning pipeline (IGAMS and BISP in R) to extract features and build predictive models for stage prediction for kidney cancer subtype by integrating DNA methylation and RNA-Seq data.
• The models were integrated using different multiple kernel learning approaches, showing an improvement of over 5% in MCC and accuracy compared to the best-performing models built over the individual modalities.
• Identified and studied the genes responsible for stage progression and investigated the relationship with the observed methylation profile.

Aug 2017 – Dec 2017 · 4 mos

• Game Design & Engineering

May 2017 – Jul 2017 · 2 mos · Heidelberg, Baden-Württemberg, Germany

• Designed a parallelized, automated pipeline (Python) that investigated more than 39K Chip-Seq datasets for their enrichment at specific genomic loci. Novel proteins were found through the analysis which were also experimentally verified.
• Contributed to interactive visualization library LinkedCharts (Javascript, D3) designed for analyzing and drawing insights from different high-throughput genomic data.

Jan 2017 – Apr 2017 · 3 mos

• Electromagnetism & Optics

Aug 2015 – Dec 2015 · 4 mos

• Science 1