If you know of open roles for which expertise in ML, simulation, or modelling as applied to drug/protein/materials design is sought, please feel free to reach out -- any pointer will be much appreciated!
As an opportunity for change, I’m also open to roles that may stretch my skills and expertise. Due to personal constraints, I’ll be looking for positions primarily in the SF Bay and Boston areas.
My colleagues also have considerable experience in ML and drug discovery, and I can’t say enough good things about them. Again, feel free to reach out if aware of potentially suitable positions.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
2/5 In molecular design, structure-property relationships are often qualitatively or quantitatively analyzed to guide the navigation of chemical space, with rougher landscapes generally expected to pose tougher optimization challenges.
3/5 To quantitatively capture the roughness of molecular property landscapes, we propose an index that is loosely inspired by the concept of fractal dimension.
1/7 In this preprint, @cwcoley and myself try to expand the domain of applicability of graph-based representations and models from well-defined molecules to materials that are ensembles of similar molecules, like polymers. arxiv.org/abs/2205.08619
2/7 We focus in particular on copolymers. The core idea is simple: describe a molecular ensemble by its average graph structure using weighted edges. For polymers, this captures the average repeating unit. We then used an MPNN with messages weighted accordingly.
3/7 We built a computational dataset with ~40k copolymers with varying monomer identities, stoichiometries, and chain architectures. On random splits, this representation returned RMSEs ~5 times lower than a baseline MPNN, and ~3 lower than the next-best model.