Data Scientist - Biomedical Knowledge Graph - REMOTE

Rancho Biosciences - Boston, MA

Full-time

Remote - Boston, MA

Professional, Scientific, and Technical Services

About the position

The Biomedical Knowledge Graph Data Scientist at Rancho BioSciences will focus on developing and optimizing pipelines that utilize knowledge graph embeddings (KGE) for drug discovery. This role involves implementing KGE models, assessing their predictive performance, and ensuring their relevance in practical biomedical contexts. The position is remote and offers opportunities for professional and personal growth within a fast-paced work environment.

Responsibilities

• Design and build scalable data pipelines to implement and execute KGE models on drug discovery knowledge graphs.
• Perform thorough assessments of KGE models to gauge their efficacy in biomedical contexts.
• Create and run tests to explore how various training methodologies and settings affect model performance.
• Apply advanced techniques for optimizing hyperparameters to enhance model precision and adaptability.
• Work in tandem with interdisciplinary groups to ensure KGEs are relevant to practical drug discovery and meet standards for equitable evaluation and replicability.
• Compile and present insights and suggestions for enhancing KGE model assessment practices.

Requirements

• Master's or Ph.D. in Data Science, Bioinformatics, Computer Science, or a related field.
• Demonstrated success in developing and overseeing data pipelines and managing extensive datasets.
• Advanced Python programming abilities, particularly with PyTorch and associated libraries like PyG and PyKEEN.
• Practical knowledge of knowledge graphs, machine learning techniques, and graph embedding models, including their real-world applications.
• Acquaintance with biomedical knowledge graph platforms, including but not limited to Disqover, PrimeKG, Hetionet, or BioKG.
• Proven proficiency in fine-tuning parameters and optimizing models, with specific experience in Bayesian optimization methods.
• Superior analytical capabilities and the skill to effectively convey complex concepts to diverse audiences.

Nice-to-haves

• Experience with biomedical datasets or in the drug discovery field.
• Familiarity with computational biology and systems pharmacology principles.
• Strong understanding of evaluation metrics and best practices for ensuring model reproducibility in scientific research.

Benefits

• Competitive salary
• Comprehensive benefits package