Advisor - Antibody Developability Validation & Benchmarking

Lilly•Indianapolis, IN

1d•$166,500 - $266,200•Hybrid

About The Position

The Advisor/Senior Advisor - Antibody Developability Validation & Benchmarking plays an essential role in establishing whether TuneLab's federated antibody models can be trusted to triage real candidates. The person in this seat must understand, at depth, how antibodies are characterized, what makes a sequence developable or not, and how predictions from a federated model translate into go/no-go decisions in a discovery pipeline. This is a validation-led role that contributes to model design choices. The person will partner closely with antibody modeling scientists on architecture, feature design, and uncertainty quantification — not just downstream of them.

Requirements

PhD in Computational Biology, Bioinformatics, Computational Chemistry, Computer Science, Statistics, or related field from an accredited college or university
Minimum of 4 years of post-PhD experience working with antibody discovery, engineering, or developability data in a biopharmaceutical or related setting
Demonstrated experience analyzing or modeling data from antibody developability assays (e.g., HIC, AC-SINS, nanoDSF, polyspecificity panels, viscosity, chemical liabilities), evidenced by publications, project work, or thesis
Hands-on experience with antibody numbering tools (ANARCI or equivalent) and working knowledge of Kabat, Chothia, and IMGT numbering schemes
Demonstrated experience designing ML validation protocols for biological sequence data, including sequence-similarity-aware splits and held-out test design

Nice To Haves

Experience fine-tuning protein or antibody language models (e.g., ESM-2, AbLang, IgBERT, AntiBERTa) for property prediction tasks, including self-supervised pretraining on OAS and fine-tuning strategies for low-data developability endpoints
Working knowledge of sequence liability motifs (Asp isomerization, Met oxidation, deamidation, glycosylation sites in CDRs)
Strong foundation in experimental design, statistical validation, and hypothesis testing
Proficiency in data engineering, pipeline development, and automation
Experience with NVIDIA FLARE or comparable federated learning frameworks (Flower, OpenFL, PySyft)
Working knowledge of antibody structure prediction tools (AlphaFold-Multimer, IgFold, ABodyBuilder) and how their outputs feed downstream developability models
Familiarity with public antibody resources — SAbDab, OAS, TAP, Jain panel, FLAb
Understanding of the manufacturability funnel from discovery through CLD and formulation, and which developability properties gate which stage
Knowledge of regulatory considerations for AI/ML in pharmaceutical development
Experience with uncertainty quantification methods (conformal prediction, Bayesian approaches, ensemble disagreement) and calibration assessment
Proficiency in PyTorch and the modern ML ecosystem (Hugging Face, scikit-learn, RDKit)
Experience with experiment tracking and model registry tools (MLflow, Weights & Biases)
Publications on antibody developability prediction, model validation, benchmarking, or reproducibility
Exceptional attention to detail and commitment to scientific rigor
Strong technical writing skills for partner-facing model cards and validation reports
Portfolio mindset balancing rigorous validation with rapid deployment for partner value

Responsibilities

Build the canonical benchmark suite covering the full developability portfolio — aggregation propensity (AC-SINS, SMAC, CIC), thermal stability (nanoDSF/DSF), polyspecificity (BVP-ELISA, Heparin RT, PSR), self-interaction, viscosity, chemical liabilities (deamidation, isomerization, oxidation, N-glycosylation in CDRs), and immunogenicity surrogates. Define which endpoints are evaluated jointly versus independently and how multi-endpoint reliability rolls up to a triage decision.
Architect privacy-preserving protocols for constructing representative test sets across distributed partner datasets, with splitting strategies appropriate to antibody data — germline-based, CDR-similarity-based, and clonotype-based splits that genuinely test generalization rather than near-duplicate memorization. Account for the structural asymmetry of antibody data (many sequences with shallow characterization, few sequences with deep characterization) when designing held-out evaluation sets.
Systematically benchmark federated antibody models against established external resources — SAbDab, OAS, TAP, the Jain et al. clinical-stage antibody panel, FLAb, and equivalent emerging datasets — to characterize generalization gaps and quantify where federated training delivers measurable lift over public-only baselines.
Develop validation strategies that assess model generalization across modalities and formats relevant to antibody developability — IgG vs. bispecific vs. fragment formats, different expression systems, different assay protocols across partners — while respecting partner data boundaries.
Implement temporal-split and sequence-similarity-aware validation protocols that simulate prospective deployment, detect concept drift as partner data accumulates, and surface systematic failure modes across CDR length distributions, germline families, and physicochemical regimes.
Work alongside antibody modeling scientists on architectural and feature choices that have direct validation implications — uncertainty quantification approaches, calibration strategies, structure-aware vs. sequence-only representations, and how predictions from different endpoints should be combined or kept independent.
Design statistically powered validation studies that account for multiple testing across endpoints, hierarchical structure in antibody data (sequences clustered by germline, project, partner), and non-independent observations. Provide honest confidence intervals on reported model performance.
Build robust MLOps pipelines ensuring complete reproducibility of federated experiments, including versioning of data snapshots, model checkpoints, and hyperparameter configurations.
Develop comprehensive performance profiling across germline families, CDR length regimes, framework variants, and property ranges, identifying systematic biases and failure modes that should be communicated to partners.
Collaborate with engineering teams to integrate validation frameworks with the TuneLab federated learning platform built on NVIDIA FLARE, ensuring scalable and automated testing across the partner network.

Benefits

company bonus
company-sponsored 401(k)
pension
vacation benefits
medical benefits
dental benefits
vision benefits
prescription drug benefits
flexible benefits (e.g., healthcare and/or dependent day care flexible spending accounts)
life insurance
death benefits
certain time off and leave of absence benefits
well-being benefits (e.g., employee assistance program, fitness benefits, and employee clubs and activities)