iiinova — where your digital self keeps evolving

Reposted by Maria A. Castellanos

Achira @achira-ai.bsky.social · 24/10/2025

Achira is growing! We’re looking for talented software engineers, ML research engineers, and AI/ML scientists to join our team in building foundation simulation models to power the future of drug discovery. Apply at achira.ai

achira.ai

Achira

Building foundation simulation models for drug discovery

1136

Reposted by Maria A. Castellanos

MSKEducation @mskeducation.bsky.social · 18/09/2025

Kicking off the #MSKPostdoc Research Symposium: @mariacastel.bsky.social, Andrew S. Gordon MERIT Fellow from the @jchodera.bsky.social lab, presents a computational pipeline to design antivirals that target conserved protein sites - arming us against future pandemics. 🧬💻🦠 #NPAW2025 #VirusResearch

061

Reposted by Maria A. Castellanos

Sukrit Singh @sukritsingh92.bsky.social · 14/03/2025

Happy Holi, from the @jchodera.bsky.social lab!!

Five people, covered in bright Holi powder, stand arm-in-arm in front of a waterfront railing with a cityscape in the background. They’re all smiling and wearing colorful, powder-stained clothes.

0282

Reposted by Maria A. Castellanos

John Chodera @jchodera.bsky.social · 19/02/2025

Everything is chaos, but I wanted to share some awesome recent science from the lab that hints at where the future of biomolecular simulation is headed: Foundation simulation models that can be fine-tuned to experimental free energy data to produce systematically more accurate predictions.

Figure 1 from arXiv preprint https://doi.org/10.1101/2025.01.06.631610

Fig. 1 Espaloma is an end-to-end differentiable molecular mechanics parameter assignment scheme for arbitrary organic molecules. Espaloma (extensible surrogate potential optimized by message-passing) is a modular approach for directly computing molecular mechanics force field parameters FFF from a chemical graph G such as a small molecule or biopolymer via a process that is fully differentiable in the model parameters FNN. In Stage 1, a graph neural network is used to generate continuous latent atom embeddings describing local chemical environments from the chemical graph. In Stage 2, these atom embeddings are transformed into feature vectors that preserve appropriate symmetries for atom, bond, angle, and proper/improper torsion inference via Janossy pooling.54 In Stage 3, molecular mechanics parameters are directly predicted from these feature vectors using feed-forward neural networks. This parameter assignment process is performed once per molecular species, allowing the potential energy to be rapidly computed using standard molecular mechanics or molecular dynamics frameworks thereafter. The collection of parameters FNN describing the espaloma model can be considered as the equivalent complete specification of a traditional molecular mechanics force field such as GAFF38,39/AM1-BCC55,56 in that it encodes the equivalent of traditional typing rules, parameter assignment tables, and even partial charge models. Reproduced from ref. 49 with permission from the Royal Society of Chemistry.

310530

Reposted by Maria A. Castellanos

Polaris @polarishub.io · 03/12/2024

🦠 We’re excited to announce our first competition in partnership with @asapdiscovery.bsky.social and @omsf.io! Test your skills across three sub-challenges revolving around SARS-CoV-2 and MERS-CoV Mpro🧵 Full details: polarishub.io/competitions Blog: polarishub.io/blog/antivir...

14316