Postdoc: AI-Driven Predictive Biology for Genomics

Lawrence Berkeley National Laboratory; Berkeley (CA); On-site; USD 92,000 - 103,000; Full time

Benefits offered by this job

Exceptional health benefits

Generous paid time off

Inclusive culture

Job summary

Lawrence Berkeley National Laboratory in Berkeley, CA, seeks a Postdoctoral Fellow to join the US DOE's KBase team. You will adapt and augment genome and protein foundation models for biological function prediction, using KBase data and benchmarks, and publish findings.

The role emphasizes AI/ML, interdisciplinary collaboration, and open science, with a start date around October 1, 2026 and a full-time, two-year appointment.

Qualifications

• PhD in Computational Biology, Bioinformatics, ML, CS, Statistics, or related field.
• Strong AI/ML background for predictive biology.
• Experience adapting pretrained DL models to scientific tasks (fine-tuning, parameter-efficient methods).
• Experience applying DL to biological sequences or omics data.
• Rigorous model evaluation and experimental design.
• Strong Python programming and PyTorch proficiency.
• Excellent communication and publication record.
• Interdisciplinary collaboration experience.

Responsibilities

• Adapt and augment genome and protein foundation models for function prediction.
• Augment models with KBase functional signals and context.
• Develop benchmarks tying model performance to biological tasks with defined splits and baselines.
• Evaluate model calibration and uncertainty; analyze failure modes.
• Deploy validated models and adapters within KBase capabilities.
• Publish research findings, methods, benchmarks, and datasets; release evaluation code.

Skills

Python

PyTorch

Machine Learning

Computational Biology

Data analysis

Research communication

Education

PhD in Computational Biology / related field

Job description

Lawrence Berkeley National Laboratory in Berkeley, CA, seeks a Postdoctoral Fellow to join the US DOE's KBase team. You will adapt and augment genome and protein foundation models for biological function prediction, using KBase data and benchmarks, and publish findings.

The role emphasizes AI/ML, interdisciplinary collaboration, and open science, with a start date around October 1, 2026 and a full-time, two-year appointment.