Thanks to @pubs.acs.org for accepting #MyACSCover featuring our work.

Siponimod (red) binds strongly to both sphingosine-1-phosphate (S1P) receptor 1 (green, active) and receptor 2 (blue, inactive), but only S1PR1 undergoes the conformational changes needed for activation.

⚙️ Using long-timescale molecular dynamics (#MD), ensemble analysis & #GPCR transmission-switch reasoning, we explain a clinically relevant selectivity and safety phenotype.

✨ Unearthing the Gold Dust of Mechanistic Insight With Deep Tech✨

🧪 We show this in our recently published ACS
@JCIM_JCTC paper, elucidating the mechanistic basis by which Siponimod, a drug for #multiplesclerosis (MS), avoids off-target activation.

pubs.acs.org/doi/10.1021/...

Meet @madhuram.bsky.social and Raj Kumar Chakraborty at the #Peptides & Complex Generics Symposium to see how Agentic & Generative #AI along with molecular simulations are accelerating Target elucidation, #LNP engineering and more for peptides & other complex biological therapeutics.

🚀Real-World Applications of #AgenticAI Accelerating #Biopharma R&D and Care

👉 Join us at AIPM-India 2025 in IIT Madras to see in action, Igniva™ AI agents, built on 5+ years of global biopharma R&D collaborations by 100+ researchers at Aganitha.

We're #hiring for multiple roles in our Computational Chemistry team at Aganitha (Hyderabad, India).

Meet us at the 6th International #Genomics Analysis & Technology Conference (#GATC) to learn about:

🧬 Causal inference on biobank-scale genomics & plasma proteomics for Biomarker & Target identification

🧩 Deciphering role of non-coding variants in disease

🤖 AI-powered Variant Effect Prediction

Meet us at "Recent Advances in Modeling Rare Events (RARE) 2025" conference this week, to see how #AI and enhanced #MD are transforming #biopharma R&D.

🎤 Join @madhuram.bsky.social's talk on #AI driven #PROTAC design at #ProUPS-25.

#AI #AI4Science #GenAI #AgenticAI #PMWC25

📢 Join us at #PMWC25 to talk about AI in BioPharma R&D

Vikram Duvvoori, CEO of Aganitha.ai, will be presenting case studies showing how #GenerativeAl & in silico modeling are mining insights from deep science to accelerate #BioPharma R&D. Event details in the poster below. See you there.

🧬🧪🧶 For penetrants, the N-terminal of the LY6A receptor engages extensively with trimeric #AAV9 capsid protein. Conversely, for non-penetrants, the trimeric AAV9 interacts only with loop VIII from one of the three chains in the trimer.

🧬🧪🧶 Molecular dynamics (#MD) simulations showed 4x improvement in binding to LY6A receptors that were previously shown to be targets of #BBB penetrant variants in mice models used by DE experiments.

🧬🧪🧶 Here are the changes optimizer made to maximize the penetration for brain endothelial cells, starting from TLQLPFK, a top 7-mer insert found by M-CREATE to yield a good enrichment score for multiple brain cell types. #AI

🧬🧪🧶 Loss function for the optimizer combined predicted fitness score along with a penalty for straying outside the embeddings sub-space covered by the training dataset. #AI

🧬🧪🧶 We designed a more direct approach: Autograd based maximisation of protein fitness (#AutoMaxProFit).

Using data from DE, we fine-tuned a #Transformer based protein language model to predict fitness scores, performed gradient ascent in the embedding space & decoded optimal seq. #AI