Principal Computational Biologist

Gordian Biotechnology•South San Francisco, CA

5d•$170,000 - $230,000•Hybrid

About The Position

Gordian has generated in vivo perturbation data spanning over 1000 targets across Obesity, Heart Failure, Pulmonary Fibrosis, Osteoarthritis, MASH, and Chronic Kidney Disease (CKD), while partnering with external collaborators to build the most comprehensive knowledge graph of disease-relevant, translatable therapeutics in this space. As a Principal Computational Biologist, you go beyond executing individual screen analyses: you pioneer how Gordian strategizes around perturbation data. You help define the methods that get standardized across the computational team, and keep up with the blistering pace of AI in the space, integrating the best tools into our current and future workflows. You partner directly and autonomously with disease-area experts on forward screen planning, with minimal oversight, and you're expected to spot opportunities the rest of the team hasn't yet seen. Target prioritization and validation are core to this role. You take screen results beyond hit-calling, building the analytical case for which targets are most likely to be both mechanistically valid and therapeutically actionable, and partnering with disease experts to design the follow-on validation that tests those calls. This role carries a strong need for familiarity in the applied ML and perturbation response prediction space given single-cell omics data, spanning perturbation-response models, trajectory inference, and representation learning. You bring real depth here, including applying methods you're already expert into a biological context that's new to you, as well as integrating the new methods being developed daily by the community. You also have experience integrating orthogonal data modalities, such as proteomics, human genetics, biomarkers, and public resources like GTEx or UK Biobank, with our single-cell screen data to strengthen translatability and sharpen target prioritization decisions. Beyond your own projects, you help define and standardize the statistical frameworks, controls, and QC criteria the broader computational team relies on, and you're expected to reason quickly and well about whether a method or dataset from one disease context transfers to another. You'll also help define how Gordian deploys agentic LLM systems to build modular, semi-automated frameworks for QC, analysis, and interpretation across the team, and maintain the standards for inter-team collaborations with our wet-lab functional groups (single cell, molecular biology, and in vivo groups) to ensure conclusions are always met with the most appropriate quantitative analyses, and nuances are always communicated in the most transparent manner. You play a key role in shaping how these experiments are designed. By three months, you'll be making significant contributions to feature development and screen validation frameworks, evaluating alternative analytical approaches with an emphasis on interpretability tied to mechanism-of-action. At six months, you'll have defined strong positive and negative controls used across multiple screens, be operating independently with disease experts on forward screen planning, and be proposing concrete improvements to team-wide analysis methodology, with visible influence beyond your own individual analyses.

Requirements

Ph.D. in Bioinformatics, Computational Biology, Statistics, Computer Science, or a related quantitative field.
Deep domain expertise in disease biology.
At least 2+ years of hands-on post-graduate industry experience (postdoctoral experience alone does not establish this).
Demonstrated track record of applying computational biology to single-cell transcriptomic data in preclinical or translational settings.
Experience extending beyond hit-calling or descriptive cell-state characterization to target prioritization, mechanism-of-action investigation, biomarker discovery, or validation, ideally in work that progressed toward or was explicitly designed to support clinical translation.
Highly capable in R and/or Python.
Ability to work fluently with major single-cell ecosystems such as Seurat, Scanpy, and related tools.
Ability to select and adapt methods based on the biological question rather than being tied to a particular computational framework.
Experience leveraging large public or external biological resources such as GTEx, UK Biobank and other human genetics/GWAS resources, or other disease-relevant omics datasets, and using them to provide context, validate hypotheses, prioritize targets, identify biomarkers, or inform a biological or translational decision.
Genuine drive to develop or substantially extend computational methods when existing approaches are insufficient, with at least one peer-reviewed publication or preprint in which you were a major contributor demonstrating this capability.
Ability to recognize when an analytical problem is genuinely novel, formulate an appropriate computational strategy, and determine how to establish whether the resulting approach is actually useful.
Track record of standardizing or scaling analytical workflows across a team, not just for your own projects.
Demonstrated experience integrating multiple layers of biological evidence to answer a scientific or translational question (e.g., single-cell data with animal phenotypes or assay metadata like histology, proteomics, etc.).
Comfortable moving from molecular and cellular observations to tissue-level phenotypes and ultimately to hypotheses about mechanisms of action with empirical support.
Strong machine learning and computational modeling background applied in real contexts, deep enough to bring rigor to a new biological domain even without prior exposure.
Excellent interdisciplinary communicator, comfortable being the person a disease lead relies on to make judgment calls independently.
Proactive about asking the right clarifying questions when extending expertise into a new context.
Experience mentoring junior computational candidates (RA/graduate level).
Experience working in a group.
Track record of success in high-agency work, consistently creating momentum rather than waiting for direction.
Desire to do best work alongside exceptional teammates and be energized by environments where people push each other to think more clearly, work at a higher standard, and grow.
Desire to play a key role in an early-stage startup screening new targets for intractable diseases of aging.

Nice To Haves

Direct prior work in cardio-renal-metabolic biology and relevant tissues (heart, kidney, adipose, liver).
Closely related experience (e.g., UK Biobank and other relevant GWAS catalogues, QTL analyses coupled with single-cell data) demonstrating transferable fluency in gaining deeper insights into MOA and target prioritization.
Experience with pooled perturbation and CRISPR screening data specifically.
Experience with spatial transcriptomics data (10X Xenium, Visium, etc).
Experience with preclinical models for validation with functional readouts (e.g., human explants, organoids, etc.).

Responsibilities

Pioneer how Gordian strategizes around perturbation data.
Define methods that get standardized across the computational team.
Integrate the best AI tools into current and future workflows.
Partner directly and autonomously with disease-area experts on forward screen planning with minimal oversight.
Spot opportunities the rest of the team hasn't yet seen.
Build the analytical case for which targets are most likely to be both mechanistically valid and therapeutically actionable.
Partner with disease experts to design follow-on validation that tests those calls.
Apply methods expert in a biological context that's new.
Integrate new methods being developed by the community.
Integrate orthogonal data modalities (proteomics, human genetics, biomarkers, public resources like GTEx or UK Biobank) with single-cell screen data.
Help define and standardize statistical frameworks, controls, and QC criteria for the broader computational team.
Reason quickly and well about whether a method or dataset from one disease context transfers to another.
Help define how Gordian deploys agentic LLM systems to build modular, semi-automated frameworks for QC, analysis, and interpretation.
Maintain standards for inter-team collaborations with wet-lab functional groups.
Ensure conclusions are met with the most appropriate quantitative analyses.
Communicate nuances in the most transparent manner.
Play a key role in shaping how experiments are designed.
Make significant contributions to feature development and screen validation frameworks.
Evaluate alternative analytical approaches with an emphasis on interpretability tied to mechanism-of-action.
Define strong positive and negative controls used across multiple screens.
Operate independently with disease experts on forward screen planning.
Propose concrete improvements to team-wide analysis methodology.