POSITION SUMMARY:
Natera is seeking a highly skilled Machine Learning Scientist to contribute to cutting-edge research and development within our Artificial Intelligence and Machine Learning team, focused on molecular therapeutics and biomarker discovery. The successful candidate will play a key role in implementing and expanding state-of-the-art deep learning models and foundation models within the biomedicine space, leveraging extensive hands-on experience and a proven track record. You will also work closely with researchers and medical professionals to interpret model results and revolutionize medical diagnostics and treatments.
RESPONSIBILITIES:
- Design, implement, and evaluate deep learning models, such as deep sequence and biomedical foundation models, with scientific rigor for applications in therapeutics, prognostics, and diagnostics.
- Independently explore and stay current with the latest advancements in AI/ML and biomedical literature, applying cutting-edge AI/ML approaches to solve complex problems in molecular diagnostics and therapeutics.
- Interpret machine learning models by explainable AI methods (e.g. GradCAM, Shap) on diverse data modalities, such as genomics, transcriptomics and imaging, generating insights to enhance model transparency and trust.
- Document code and model by maintaining well-structured, readable, and reproducible workflows. Use version control systems (e.g. Git) and reproducible demos (e.g. Jupyter notebooks) to ensure consistency across projects.
- Communicate scientific results effectively through internal meetings, external conferences, and peer-reviewed publications, translating complex technical concepts for both technical and non-technical audiences.
- Collaborate with domain experts such as medical professionals, bioinformaticians, and statisticians to guide and refine R&D efforts in AI-driven medical diagnostics and therapeutics.
QUALIFICATIONS:
-
PhD in Computer Science, Statistics, Bioinformatics, or a related quantitative field
KNOWLEDGE, SKILLS, AND ABILITIES:
- Proven experience with deep sequence models, biomolecular foundation models, graph neural networks, modality fusion, representation learning, generative models, and self-supervised learning techniques; and a demonstrated track record of machine learning applications to genomics, proteomics, and/or multi-omics big data.
- Highly proficient in PyTorch, lightning, tensorboard, captum, scikit-learn, transformers, and Huggingface APIs.
- Experience with SQL databases (snowflake and mysql-server).
- Experience in Git, jupyter notebook, and advanced visualizations (seaborn, plotly)
- Experience with AWS, single and multiple GPU-accelerated environments
- Ability to clearly summarize and communicate results in oral and written form.
- High scientific rigor and eagerness to teach and learn about new machine learning methods and biology.
- Experience with oncology or immunology is a plus
- Experience with medical imaging and joint modeling of cancer genomics and imaging is a plus
