With all of the TCR/pMHC binder papers that have recently come out (Baker Lab, BIKEs from Jenkins lab, etc.) and general interest in de novo peptides and biosecurity, ive been interested in how you screen for immunogenicity in silico for these models.
Self-driving labs & autonomous science are super exciting, but feels like we will hit a bottleneck in automation as cognition improves. There's an inherent asymmetry - the most valuable experiments to automate (like in vivo work) are often the least automatable
Incredibly excited for this to be announced! MD simulation has been too long constrained by rules tailored to compute from the ‘70s. Patchwork solutions (incl some that I worked on) won’t cut it — we need a new paradigm
We're launching ACHIRA, a newco DIMENSION cofounded alongside @jchodera & @tkaraletsos. We're building foundation simulation models for the smallest, atomic-level resolution units of the universe. Assembling here: achira.ai
1/ Some pre-thanksgiving musings! I've been thinking about new approaches to challenging drug targets like MYC - particularly on targeting intrinsically disordered proteins (IDPs). These proteins challenge our traditional understanding of structure-function relationships. 🧵
NEWS: Today we're launching Dimension II, $500M of fresh capital to support and partner with the world's leading entrepreneurs at the cutting most edges of life sciences and technology!!
Enjoyed bringing together a small group of brilliant and motivated students and early-career scientists at the first @_DimensionCap × @openbio_lab AI for Biology dinner.
An evening of brainstorming about the future of AI for life science with new friends and great food. 🤖🧬 1/4
As small molecule drugs become larger, biologics become smaller, and peptide therapeutics gain adoption, how will it act as a forcing function for new representations in ML applied to molecular design?
Do we default to atoms/electrons vis-a-vis neural network potentials?
Does
Thank you so much to @Lux_Capital, @wolfejosh & @velvetatom for hosting the Princeton Biotech Group today! Thank you as well to everyone who attended our happy hour. We all left feeling incredibly inspired to delve further into the world of biotech🧬
Plus there's a long tail of valuable but infrequent research tasks that don't justify building integrated workflows. I think we need both generalizable robotics for labs and a universal software protocol for tools (i.e MCP for lab instruments) to fully do autonomous science.
Thank you as well to @abscibio, @joshim5, @rayrah_, @HarmonicDiscov, Cat Donaldson & Alexandria Launch Labs for talking to us about the amazing work they are doing and showing us around! Big thanks to @MichaelRetchin for helping plan this