Experimental.
Goal: Interactive derivation of biological facts using an inference engine.
See pinky-pipeline for details.
Initial data includes the following:
- Open Targets Platform evidence from ChEMBL (gene-drug-clinicalTrial-disease)
- Open Targets Genetics locus to gene scores (gene-locus-study-trait)
- StringDB (protein-protein)
- Ensembl (gene annotation)
- HGNC (gene annotation)
- UniProt (protein annotation)
See pinky-api for details.
Inference managed by clara-rules, an implementation of the Rete algorithm. Custom inference rules are Clojure functions that match on records/facts derived from the pipeline data.
TODO!