Computational Scientist

Pharmaceutical Company; South San Francisco (CA); On-site; USD 120,000 - 180,000; Full time

Benefits offered by this job

12-month contract

Cluster computing access

Job summary

Pharmaceutical Company in South San Francisco, CA seeks a Computational Scientist to advance research on large-scale multimodal perturbation datasets.

The role focuses on training and evaluating ML models to understand dataset contributions to performance, running controlled experiments, and comparing predictions with experimental results, in collaboration with experienced scientists.

Qualifications

• PhD in a computational discipline or equivalent research experience.
• Backgrounds in Computational Biology, Computational Sciences, Statistics, Machine Learning, or related fields.
• Hands-on experience training and evaluating deep learning models using PyTorch.
• Experience designing and executing controlled train/test split experiments.
• Experience working with SLURM-based computing clusters or HPC environments.
• Strong applied statistics and analytical problem-solving skills.
• Proficiency in scientific Python, including NumPy and Pandas.
• Experience with reproducible and well-documented computational analysis practices.

Responsibilities

• Train and evaluate deep learning models using large-scale multimodal perturbation datasets.
• Investigate how different subsets and components of datasets contribute to model learning and performance.
• Design and execute controlled train/test split experiments to assess model performance.
• Perform computational experiments using PyTorch and SLURM-based HPC environments.
• Apply statistical methods to evaluate model performance, interpret results, and identify meaningful patterns.
• Compare model predictions and outputs against experimental data.
• Develop reproducible, well-documented scientific analyses using Python.
• Collaborate with research scientists and project leads to deliver defined computational research objectives.

Skills

Deep learning

Python programming

Experiment design

Statistics

HPC computing

Education

PhD in computational discipline

Tools

PyTorch

SLURM

NumPy

Pandas

Job description

We are seeking a Computational Scientist to support research involving large-scale multimodal perturbation datasets. The role will focus on training and evaluating machine learning models to understand how different portions of complex datasets contribute to model performance and learning.

The selected candidate will analyze multimodal experimental data, conduct controlled computational experiments, and evaluate model predictions and outputs against experimental results. This position offers an opportunity to contribute to a high-impact computational research initiative while collaborating with experienced scientists and research teams.

Key Responsibilities

• Train and evaluate deep learning models using large-scale multimodal perturbation datasets.
• Investigate how different subsets and components of datasets contribute to model learning and performance.
• Design and execute controlled train/test split experiments to assess model performance.
• Perform computational experiments using PyTorch and SLURM-based high-performance computing environments.
• Apply statistical methods to evaluate model performance, interpret results, and identify meaningful patterns.
• Compare model predictions and outputs against experimental data.
• Develop reproducible, well-documented scientific analyses using Python.
• Collaborate with research scientists and project leads to deliver defined computational research objectives.

Required Qualifications

• PhD in a computational discipline or equivalent research experience.
• Preferred academic backgrounds include Computational Biology, Computational Sciences, Statistics, Machine Learning, or related fields.
• Hands-on experience training and evaluating deep learning models using PyTorch.
• Experience designing and executing controlled train/test split experiments.
• Experience working with SLURM-based computing clusters or high-performance computing environments.
• Strong applied statistics and analytical problem-solving skills.
• Proficiency in scientific Python, including NumPy and Pandas.
• Experience with reproducible and well-documented computational analysis practices.

Preferred Qualifications

• Experience with single-cell transcriptomics.
• Experience analyzing CRISPR perturbation screens, including Perturb-seq or similar datasets.
• Familiarity with representation learning techniques.
• Familiarity with image analysis and multimodal datasets.

Contract Highlights

• Initial 12-month contract with full-time commitment.
• Ownership of a clearly defined research workstream with measurable deliverables.
• Direct collaboration with project leadership and experienced research scientists.
• Access to an established, documented analysis codebase and cluster computing resources.
• Opportunity to contribute to cutting-edge computational biology and biotechnology research.