ADC Design Knowledge Base demo
A cited, structured memory of every antibody–drug conjugate built: antibody, linker, payload, drug-to-antibody ratio, conjugation chemistry, efficacy, safety and manufacturability outcomes, searchable in plain language. 1:14
What the demo shows
An antibody-drug conjugate solution for ADC scientists that records every design and result.
Extracts construct-level facts from historical and current reports into one structured ADC registry (target, antibody format, linker, payload class, drug-to-antibody ratio and its distribution, aggregate at release, free payload, plasma stability, efficacy, organ findings and why each construct stopped), every field carrying the document, page and sentence behind it. A design question lights up the molecules it is about on the design space and returns a cited lesson; every new design proposal is scored against the whole portfolio before the board, with the precedents, the risks and the changes that worked.
Transcript
- The ADC Design Knowledge Base holds every antibody drug-conjugated team has built for the scientists designing the next one.
- Toxicity is the leading reason these drugs fail and the finding usually belongs to the chemistry.
- Fourteen constructs by ratio and payload class.
- Color is how each one ended.
- Ask a design question and the molecules behind the answer light up.
- Above six, the hydrophobic linker failed the aggregate limit.
- The PEG extended linker held 2.3%.
- Every number opens the report at its page.
- Maya proposes a second run on that chemistry.
- Against everything built before, its repeat risk is 78.
- Three changes that worked before Bring It to 21, beside constructs that cleared.
- Built on Botminds: agents, data, screens, and governance on one platform. Thank you.
Transcribed from the film’s narration. Captions are on the player.