Scientist I, Machine Learning
Foundation Medicine
Summary
The Scientist I, Machine Learning contributes to research, implementation, and validation of computational methods for FMI's internal Computational Discovery Research group. This position supports the development of machine learning algorithms and data pipelines applied to large-scale biomedical datasets including histopathology imaging, genomic, transcriptomic, and clinical outcomes data. The Scientist I works closely with other machine learning scientists, computational biologists, clinicians and software engineers to contribute to workflows that may include the investigation and identification of novel biomarker signatures, discovery of novel cancer genomics findings, support of critical data science partnerships, and improvement to FMI's operational pipelines.
Job description
Key Responsibilities
- Develop, train, and evaluate machine learning and deep learning models using large-scale structured and unstructured datasets (images, free text) to extract biological insights
- Develop data pipelines, infrastructure, and computational tools for image or genomic analyses
- Contribute to data preparation, feature engineering, model validation and performance assessment
- Apply established machine learning and statistical methods with guidance from senior team members
- Provide scientific expertise and support for internal teams and external collaborators
- Conduct novel cancer genomics research using both public and internal datasets
- Collaborate with cross-functional teams including computational biology, pathology, and engineering
- Implement reproducible analyses and contribute to codebases
- Prepare reports and presentations to communicate results in group meetings
- Present novel findings via abstracts or manuscripts
Basic Qualifications
- Bachelor's Degree in Computer Science, Bioinformatics, Computational Biology, Engineering, or other similar quantitative discipline and 3+ years of work experience in relevant field; OR
- Master's Degree in Computer Science, Bioinformatics, Computational Biology, Engineering, or other similar quantitative discipline and 2+ years of experience in relevant field
Preferred Qualifications
- Ph.D. degree in Computer Science, Bioinformatics, Computational Biology, Engineering, or other similar quantitative discipline
- Experience with deep learning (particularly foundation models, vision transformers) methods and frameworks and a strong understanding of their mathematical foundations
- Knowledge of cancer biology and cancer genomics
- Intermediate proficiency or higher in object-oriented programming with Python, Java, or C++
- Experience with traditional machine learning methods and packages (e.g. sklearn) and a strong understanding of their mathematical foundations
- Experience with distributed processing and computation (Spark, Horovod, job scheduling, etc.) for large-scale datasets
- Experience working in shared code repositories using modern version control practices (e.g., Git, pull requests, code review)
- Familiarity with the ML development life cycle or MLOps
- Experience with histopathology analysis
- Familiarity with using cloud compute providers (AWS, GCP, etc.)
- Previous authorship/co-authorship of relevant work
- Strong communication and teamwork skills to work effectively in a flexible, cross-functional environment
- Understanding of HIPAA and the importance of patient data privacy
Requirements
- Bachelor's Degree in Computer Science, Bioinformatics, Computational Biology, Engineering, or other similar quantitative discipline and 3+ years of work experience in relevant field; OR
- Master's Degree in Computer Science, Bioinformatics, Computational Biology, Engineering, or other similar quantitative discipline and 2+ years of experience in relevant field
- Ph.D. degree in Computer Science, Bioinformatics, Computational Biology, Engineering, or other similar quantitative discipline
- Experience with deep learning (particularly foundation models, vision transformers) methods and frameworks and a strong understanding of their mathematical foundations
- Knowledge of cancer biology and cancer genomics
- Intermediate proficiency or higher in object-oriented programming with Python, Java, or C++
- Experience with traditional machine learning methods and packages (e.g. sklearn) and a strong understanding of their mathematical foundations
- Experience with distributed processing and computation (Spark, Horovod, job scheduling, etc.) for large-scale datasets
- Experience working in shared code repositories using modern version control practices (e.g., Git, pull requests, code review)
- Familiarity with the ML development life cycle or MLOps
- Experience with histopathology analysis
- Familiarity with using cloud compute providers (AWS, GCP, etc.)
- Previous authorship/co-authorship of relevant work
- Strong communication and teamwork skills to work effectively in a flexible, cross-functional environment
- Understanding of HIPAA and the importance of patient data privacy
Benefits
- Discretionary annual bonus
- Benefits package
Company overview
Foundation Medicine is a leading precision medicine company headquartered in Boston, Massachusetts, that specializes in developing and commercializing comprehensive genomic profiling assays for cancer diagnosis and treatment. Founded in 2010 as a spin-off from the Broad Institute, the company leverages next-generation sequencing technology to provide molecular information that helps guide personalized cancer therapy decisions. As a subsidiary of Roche since 2018, Foundation Medicine offers a portfolio of diagnostic tests including FoundationOne CDx, FoundationOne Liquid, and FoundationOne Heme, which analyze solid tumors, hematologic malignancies, and sarcomas. The company's mission is to transform cancer care by delivering high-quality biomarker testing throughout a patient's journey, from initial diagnosis to ongoing management. With over 1,000 employees and operations in Boston, North Carolina, and Germany, Foundation Medicine collaborates extensively with biopharmaceutical partners to accelerate drug development and advance cancer research through its extensive genomic database containing profiles from more than 300,000 patients.