Mithrl is building the world's first commercially available AI Co-Scientist to accelerate the discovery of new medicines by transforming biological data into insights rapidly. The Data Scientist, Knowledge Graphs will build and scale the biological knowledge layer that powers this AI Co-Scientist.
Requirements
- Experience working with biological knowledgebases, public datasets, or ontology driven systems
- Familiarity with graph data structures, relationship modeling, and knowledge graph concepts
- Experience harmonizing heterogeneous biological datasets and mapping variable IDs across sources
- Proficiency in Python and scientific computing libraries
- Ability to build ingestion pipelines for structured or semi structured biological data
- Strong understanding of metadata standards, biological ontologies, and domain logic
- Ability to translate complex biological information into structured, machine readable representations
Responsibilities
- Ingest, harmonize, and version high value public biological datasets such as CellxGene, Gemma, ARCHS4, ENCODE, GTEx, TCGA, etc.
- Ingest well maintained peer reviewed knowledgebases including OpenTargets, HPA, and similar resources
- Build automated pipelines to curate and expand relationships inside the knowledge graph
- Define and evolve schemas for node types, relationships, metadata rules, and ontology alignment
- Harmonize variable IDs and metadata fields across all imported sources to create a unified knowledge layer
- Build and maintain versioning, change tracking, and provenance systems for all data and relationships
- Develop the framework that allows users to build custom knowledge graphs from the analyses they run inside Mithrl
Other
- Strong experience in data science, bioinformatics, computational biology, or a related field
- Excellent communication skills and comfort collaborating across engineering and scientific teams
- High-energy, in-person culture
- Comprehensive PPO health coverage through Anthem (medical, dental, and vision) + 401(k) with top-tier plans
