Advisor: Luca Pinello

• Developed PerTurbo: a scalable Bayesian analysis framework for large-scale Perturb-seq screens, with a GPU-accelerated implementation of the probabilistic model using PyTorch, Pyro, and Lightning
• Co-led construction of a Nextflow pipeline for analyzing single-cell CRISPR screen data within the IGVF consortium CRISPR Focus Group
• Refined computational strategies for single-cell lineage tracing analysis using mitochondrial mutations

Advisor: Jason Buenrostro

• Piloted use of mitochondrial mutations for lineage tracing in SHARE-seq single-cell multiomic data from cultured hematopoietic stem cells

Advisor: Shamil Sunyaev

• Improved methods for prioritizing de novo coding variants from the Undiagnosed Diseases Network (UDN) cohort by integrating a new mutation rate model with existing variant annotations

Advisor: David Pellman · Co-Advisor: Cheng-Zhong Zhang

• Created in-house Snakemake pipelines for calling copy number variants, structural variants, and single nucleotide variants from single-cell whole genome sequencing data
• Formulated a statistical approach (hidden Markov model) to accurately detect haplotype-specific loss of heterozygosity in single cells with chromothripsis
• Wrote an efficient variant caller to sensitively identify single-stranded mutations from raw reads from “bottleneck” duplex sequencing experiments

Advisor: Peter Andolfatto

• Implemented a simulation framework based on the FoldX software package to perform in silico simulations of protein evolution under physicochemical constraint

Advisor: Martin Semmelhack

• Synthesized a solvatochromic fluorophore probe covalently linked to the autoinducing peptide AI-2 from Staphylococcus aureus to study bacterial quorum sensing