Strange Loop

Building a Unified Cancer Immunotherapy Data Library

The introduction of immunotherapies has revolutionized the treatment of cancer and ushered in a corresponding explosion of research into cancer, the immune system, and their interaction. In September 2018, there were 3,394 Cancer Immunotherapy agents being tested in clinical trials - a 67% increase from the number of agents being tested one year earlier. With this explosion in research comes an explosion of data and the challenge of keeping track of and unifying published and public data so that research organizations can stay on top of the field.

To address this challenge we've created CANDEL, the Cancer Data and Evidence Library, a database system that leverages Datomic to store individual molecular measurements and clinical sample metadata in a single knowledgebase. Datomic's schema is expressive, allowing us to map the variety of representations different organizations and researchers work with into a common form, as well as to capture new concepts as our knowledge of molecular and cancer biology evolves.

Data can be loaded in the system without knowledge of Datomic internals by using a configurable, data-driven ETL pipeline we developed in Clojure. Data scientists can take advantage of the full power of datalog queries directly from their existing R analysis environment. Furthermore, Datomic's immutable history ensures any analysis will be reproducible, enabling large teams to collaborate on complex analysis projects with ease and confidence.

Lacey Kitch

Parker Institute for Cancer Immunotherapy

Lacey is Director of Informatics at The Parker Institute for Cancer Immunotherapy (PICI), a nonprofit that aims to accelerate the development of Immunotherapies and get them to patients faster through enabling great research. She has been applying data science and machine learning to biological and health-related datasets for many years, including in her PhD work at Stanford, where she developed novel algorithms to analyze optical brain imaging data to learn about memory. Now, she focuses on the development of data infrastructure and its application in translational analysis at PICI. As a cancer survivor herself, she's passionate about PICI's cause, and believes that there's a ton that good technology can contribute to accelerating cures for cancer.

Ben Kamphaus

Cognitect, Inc.

Ben Kamphaus is a software developer and data scientist at Cognitect. He has a PhD in Geography from SUNY Buffalo where he got his initial exposure to interdisciplinary challenges in data infrastructure, including the use of ontologies, RDF, and other approaches to finding and expressing common data representations. These days he relies heavily on Clojure and Datomic to stay sane when moving data between systems and use contexts. In his free time he hikes, runs, climbs, writes science fiction, and produces electronic music as PatternShift.