Working on Cancer
summary
A four‑person team spends a year turning sequencing data into ranked treatment options, ingesting and normalizing 15–20 merged databases, modeling biology in a graph and running queries in a custom DSL, handling tumors with ~10,000 mutations. The pipeline uses Clojure and Neo4j, JRuby, CoffeeScript, AWS, Puppet and Go for continuous delivery with multiple daily releases. The result is a deployed ovarian‑cancer service that ranks therapies from aggregated clinical evidence and plans to improve by feeding back outcomes.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 2:45Terminology
- 7:47Sequencing
- 11:33Cancer
- 17:05Team
- 17:58Process
- 19:44Solution
- 27:13Graph Database
- 29:59Infrastructure
- 31:45Go
- 34:54Data ingestion
- 36:42JSON
- 37:23Languages
- 40:21Conclusions
related talks
11mTracking Genome Sequence Data of Cancer Patients with Docker
31mWorking smarter to fight cancer: using dunder special methods to automate
29mObject-Oriented Oncology: Making Sense of Complex Patient Journeys
05mDuckDB powering cancer genomics research
11mEnsuring Reliability in Healthcare Data Engineering
35mSequence and Concurrent Processing for Huge Genomic Data
35mFrom cells to drug responses - machine learning in cancer research
05mStartup Showcase Winner Syapse