Daniel Himmelstein, PhD
digital craftsman of the biodata revolution
About Me
My work focuses on understanding human disease through data science, amazing software, and open collaborative approaches for the digital age.
I've recently turned my attention to radiology and artificial intelligence after an early career pioneering knowledge graphs (hetnets), machine learning, drug repurposing, data science, blockchains, graph databases, and open science.
I reside with my growing family in Lebanon, New Hampshire, part of the famed Upper Valley region.
Portfolio
AI & Automation in Radiology
2025–
Streamlining radiology reporting with RadOverlay and the TI-RADS Report Builder.
Knowledge Graphs
2012–2023
Hetionet: Integrating biology into a public Neo4j database and connectivity search.
Drug Repurposing
2013–2023
Project Rephetio: Systematic integration of biomedical knowledge to prioritize drugs.
Graph algorithms
2012–2023
Algorithms for quantifying graph connectivity between nodes in heterogeneous networks. Invented the degree-weighted path count metric. Created Hetnet Connectivity Search.
Software
Ongoing
Popular GitHub repositories and open source contributions to pandas, networkx, and more.
Ontologies
2015–2025
Creating open source software and data repositories for biomedical ontologies (nxontology, MeSH).
Target-Disease Discovery
2012–2025
Data integration for prioritizing drug targets at Related Sciences.
DeSci
2014–2020
Decentralized Science initiatives including timestamping scholarly work and incentivizing peer assessment and collaboration.
Open Manuscripts
2016–2020
Manubot: a workflow and set of tools for the next generation of scholarly publishing.
Elevation & Lung Cancer
2012–2015
Epidemiological research finding negative association between elevation and lung cancer.
Experience
Chief Technology Officer
RadOverlay
Co-founded RadOverlay to create software and automation to streamline radiology reporting.
Chief Data Scientist · Head of Data Integration
Related Sciences
Helped lead the 5-person data team to create the Facets platform for prioritizing drug targets based on large scale data integration.
Postdoctoral Fellow
University of Pennsylvania
Developed search algorithms for hetnets in the Greene Lab. Created Manubot.
PhD Biological & Medical Informatics
University of California, San Francisco
Developed metapath-based heterogeneous network analysis in the Baranzini Lab.
Summer Research
Princeton University & Dartmouth College
Internships in the Troyanskaya and Moore labs.
B.S. Biometry & Statistics
Cornell University
High School
Hanover, New Hampshire
Contact
My email is daniel.himmelstein@gmail.com. For support questions or feedback on my research, please first consider public venues — such as GitHub Issues — before emailing me.