Co-Founder, CSO at @proteinqure, @UofT @UWaterloo. Peptide Therapeutics, Structure-based Drug Design, Computational Biophysics, AI for Drug Discovery, Cats
A little extra pep in our steps for a great cause. @proteinqure joined the 2026 CIBC Run for the Cure, raising over $2,650 in support of breast cancer research!
Huge thanks to everyone who ran, donated, and supported our team, "Pep(tides) in Our Steps"! π
Three weeks ago, we welcomed the local scientific community to join us at our headquarters in Kenmore Square for the inaugural lecture in the AI Proteins Seminar Series. Our speaker Neil King from @UWproteindesign delivered a superb seminar. Check it outπ https://t.co/XdBcgtRcfm
@jcisar@proteinqure Thanks Justin! We're planning to release a few killer features, and address user feedback ahead of our talk at Boulder Peptide Symposium this year. Let me know if you have any suggestions to make this more powerful for peptide design.
@andrewwhite01@NicholasRuncie Agents tell me accuracy is the same but OpenClatura is just better...
- roughly one-third the code;
- much lower pervasive complexity;
- better diagnostics and explainability;
- richer public interfaces;
- higher single-process throughput;
@andrewwhite01 But seems like a hard problem. I ran this and STOUT against 6548 amino acid SMILES that needed IUPAC names and both tools only agreed for 2183 of them (3861 both deemed correct but differ in style, agents said openclatura got ~50 of them wrong).
@BiologyAIDaily For anyone curious, nothing appears to have changed between this version and the technical report released in January: https://t.co/Q1lD6p96XN