Functional Programming for COVID 19 Drug Repurposing
summary
A reasoning engine called mediKanren implemented in Racket and miniKanren mines merged biomedical knowledge graphs (40 sources, SemMedDB ~97M edges) to propose repurposed drugs; one chain (androgens → TMPRSS2 → SARS‑CoV‑2) surfaces anti‑androgen candidates and leads to a degarelix trial with the VA. Programmatic queries return 540 hits for androgen regulators and 1,779 for two‑hub patterns; summer interns add scripts that produce spreadsheets linking edges to PubMed sentences so clinicians triage ~1,300 drug–gene leads in a day.
AI-written summary · about
chapters (24)
- 0:00Introduction
- 1:20What is my project
- 2:02What is Medicare
- 2:35Example
- 4:43Multiple Databases
- 5:18Knowledge Graphs
- 10:44Graphical Interface
- 11:18Logic Interface
- 17:47Micropattern
- 19:14Data Structure
- 20:48Code
- 22:47What have we done
- 23:07Logic Programming
- 23:50Propagators
- 24:59Expert Systems
- 25:12Truth Maintenance Systems
- 26:35Truth System Architecture
- 27:16Biohacker
- 28:24Racket
- 32:58Debugging
- 33:17Integer Programming
- 33:37Belief Maintenance System
- 34:36Causality
- 35:42Conclusion