UCSF · Arc Institute

Ashir Borah

PhD candidate at UCSF and the Arc Institute, working across functional genomics, virology, and AI-enabled research systems.

Currently exploring |

Currently

  • Building agentic systems for scientific discovery and analysis.
  • Running screens to discover protein function.
  • Reading to build new biological systems.

Writing

  1. Loading…

All writing →

About

Biology, computation, and the tools to connect them.

I'm a PhD candidate in the Biological and Medical Informatics program at UCSF, jointly mentored by Luke Gilbert and Brian Hie at the Arc Institute. My research sits at the intersection of functional genomics, virology, and computational biology — combining CRISPR-based screening, assay development, and high-dimensional data analysis to study gene regulation, host–virus interactions, and disease mechanisms.

In parallel, I build AI-enabled research systems — LLM-based councils, research agents, APIs, and multi-agent workflows that support literature synthesis, hypothesis generation, experimental planning, and scientific reasoning.

Before UCSF I spent nearly three years at the Broad Institute on the Cancer Dependency Map, using statistical modeling and machine learning on genome-scale perturbation data to discover and validate cancer targets. I studied Mathematics and Computer Science at Dickinson College, graduating Magna Cum Laude and Phi Beta Kappa.

I'm especially drawn to problems where strong experimental design, thoughtful computation, and well-built tooling reveal biology that's otherwise difficult to measure.

  • CRISPR screens
  • Functional genomics
  • Virology
  • Host–virus biology
  • Perturb-seq
  • Machine learning
  • LLM agents
  • Research tooling

Research

What I work on.

01

CRISPR functional genomics

Designing and analyzing genome-scale CRISPR screens — knockout, interference, and Perturb-seq — to study gene regulation, dissect the programs that emerge when cells are pushed off-equilibrium, and identify therapeutic vulnerabilities.

02

Virology & host–virus biology

Building assay platforms and perturbation systems for studying virus–cell interactions and the molecular logic of infection — pairing screens with high-dimensional readouts to map the host factors that shape viral life cycles.

03

AI-enabled research systems

LLM-based councils, research agents, APIs, and multi-agent workflows that support literature synthesis, hypothesis generation, experimental planning, and scientific reasoning — building tooling that compounds rather than replaces a researcher's judgment.

04

Computational ↔ experimental loops

Closing the loop between computational predictions and bench validation — cell culture, cloning, and arrayed follow-ups — so that targets discovered in silico are stress-tested against the messiness of real biology.

Publications

Selected work.

1,125 citations · h-index 10 · i10-index 10 View on Scholar →

  1. SKI complex loss renders 9p21.3-deleted or MSI-H cancers dependent on PELO
    P.C. Borck, I. Boyle, K. Jankovic, N. Bick, K. Foster, A.C. Lau, L.I. Parker-Burns, …, A.A. Borah, et al.
    Nature 2025 Cited by 22
  2. Integrated epigenetic and genetic programming of primary human T cells
    L. Goudy, A. Ha, A.A. Borah, J.M. Umhoefer, L. Chow, C. Tran, A. Winters, et al.
    Nature Biotechnology 2025 Cited by 17
  3. A systematic search for RNA structural switches across the human transcriptome
    M. Khoroshkin, D. Asarnow, S. Zhou, A. Navickas, A. Winters, J. Goudreau, …, A.A. Borah, et al.
    Nature Methods 2024 Cited by 18
  4. In vivo Perturb-seq of cancer and microenvironment cells dissects oncologic drivers and radiotherapy responses in glioblastoma
    S.J. Liu, C. Zou, J. Pak, A. Morse, D. Pang, T. Casey-Clyde, A.A. Borah, D. Wu, et al.
    Genome Biology 2024 Cited by 18
  5. Lineage-specific canonical and non-canonical activity of EZH2 in advanced prostate cancer subtypes
    V.B. Venkadakrishnan, A.G. Presser, R. Singh, M.A. Booker, N.A. Traphagen, …, A.A. Borah, et al.
    Nature Communications 2024 Cited by 66
  6. A ubiquitination cascade regulating the integrated stress response and survival in carcinomas
    L.D. Cervia, T. Shibue, A.A. Borah, B. Gaeta, L. He, L. Leung, N. Li, et al.
    Cancer Discovery 2023 Cited by 66
  7. Partial gene suppression improves identification of cancer vulnerabilities when CRISPR-Cas9 knockout is pan-lethal
    J.M. Krill-Burger, J.M. Dempster, A.A. Borah, B.R. Paolella, D.E. Root, T.R. Golub, et al.
    Genome Biology 2023 Cited by 39
  8. Sparse dictionary learning recovers pleiotropy from human cell fitness screens
    J. Pan, J.J. Kwon, J.A. Talamas, A.A. Borah, F. Vazquez, J.S. Boehm, et al.
    Cell Systems 2022 Cited by 43
  9. Microenvironment drives cell state, plasticity, and drug response in pancreatic cancer
    S. Raghavan, P.S. Winter, A.W. Navia, H.L. Williams, A. DenAdel, K.E. Lowder, …, A.A. Borah, et al.
    Cell 2021 Cited by 611
  10. Repeat expansions confer WRN dependence in microsatellite-unstable cancers
    N. van Wietmarschen, S. Sridharan, W.J. Nathan, A. Tubbs, E.M. Chan, …, A.A. Borah, et al.
    Nature 2020 Cited by 204

See the full list on Google Scholar →

Experience

Education & positions.

  1. Arc Institute

    Jun 2023 — Present
    PhD Researcher · Palo Alto, CA
    • Designing experimental and computational methods for functional genomics and virology.
    • Building CRISPR-based screening and perturbation systems to study gene regulation, host–virus biology, and disease mechanisms.
    • Developing assay platforms and analysis workflows for high-dimensional genomic datasets.
    • Building AI-enabled scientific tooling — LLM councils, APIs, and multi-agent workflows for literature synthesis, hypothesis generation, and experimental planning.

  2. University of California, San Francisco

    Sep 2022 — Present
    PhD Candidate · Biological and Medical Informatics · San Francisco, CA
    • PhD research in functional genomics, virology, and computational biology, in collaboration with the Arc Institute.
    • Developing computational and experimental approaches for studying gene regulation, perturbation biology, and disease mechanisms.
    • Analyzing large-scale biological datasets with statistical, machine learning, and high-dimensional methods.
    • Led an R bootcamp for postdocs and clinicians — 100+ trainees.

  3. Broad Institute of MIT and Harvard

    Jul 2019 — Apr 2022
    Computational Associate I & II · Cancer Data Science · Dependency Map Project · Cambridge, MA
    • Validated candidate cancer targets identified through genome-scale CRISPR knockout screens.
    • Applied machine learning and statistical methods to identify vulnerabilities in cancer cell lines, with a focus on GI cancers.
    • Prioritized candidate dependencies for downstream therapeutic investigation.
    • Contributed to an open-source ML platform modeling genetic-knockout dependency profiles across >100,000 genomic features.

  4. Uliza

    Sep — Dec 2017
    Machine Learning Developer
    • Built ML-driven systems to support crowdsourced workflow automation.
    • Helped set up core technical infrastructure — server management, database optimization, and disaster recovery.

  5. Dickinson College

    2015 — 2019
    B.S. Mathematics & Computer Science · Magna Cum Laude · Phi Beta Kappa · Carlisle, PA
    • Student research project on secure CoAP-DTLS systems for resource-constrained devices.
    • Best Poster, All-College Science Symposium (2019).
    • Teaching assistant for Introduction to Java and Data Structures; tutoring and residential leadership roles.

Selected honors

  • Spot Award (×2), Broad Institute · 2020, 2021
  • Phi Beta Kappa Honor Society · 2019
  • Best Poster, All-College Science Symposium · 2019
  • Richard Howland Memorial Scholarship · sole CS recipient · 2018
  • Pi Mu Epsilon (Mathematics) & Upsilon Pi Epsilon (CS) · 2018
  • Jane Hill Prize in Computer Science · 2016
  • Torchbearer Award, Bhumi · 6 of 8,000 volunteers · 2015

Contact

Get in touch.

Happy to talk about CRISPR screens, virology, AI tooling for science, or graduate life at UCSF and the Arc Institute. The fastest route is email.

email me

Languages English (native) · Hindi (professional) · Assamese (conversational)