Drug discovery
Targeting and PredictionMaturity: concept
Protein/biological target identification and interaction prediction.
concept — An idea we find credible. Nothing has been built or measured.
What this is
The problem
Target identification asks which biological mechanism to intervene in at all. It is the earliest decision in the pipeline and the most expensive to get wrong — a poor target invalidates everything downstream of it.
Where the current approach strains
Evidence comes from genomics, expression data, literature and clinical observation, in different formats at different confidence levels. Integration is largely manual, and negative results are systematically under-published.
What we are exploring
Integration across those evidence types with explicit confidence propagation, so a target's support can be inspected rather than asserted.
What would have to be true
Retrospective recovery of targets now known to be valid, using only evidence available before they were validated. That is a demanding test, and the right one.
Where it applies
Related
Accelerated Drug Screening
Virtual screening has to be cheap enough to run across a whole library and accurate enough not to throw away good molecules. Docking scores track binding affinity loosely, so shortlists carry false positives you can measure and false negatives nobody ever finds out about. We are working on scoring for the middle of the funnel, tested by whether it ranks known binders highly without being told.
Molecular Optimization
Conformation, energy, binding and molecular-property optimisation.
Targeted Drug Design
Binding affinity, selectivity and off-target risk concepts for targeted compounds.