@adampecina.bsky.social
Computational Chemist | Researcher at IOCB Prague 🇨🇿 I specialize in quantum chemistry and computational drug design. Alumnus of Charles University & Palacký University. Former postdoc at the Italian Institute of Technology & Marie Curie (IF) Fellow.
Leaving Boston today after a great visit where we gave a hands-on training in #SQM affinity predictions. Now heading north to Maine, where I have the honor of chairing the GRS in #CADD & presenting our latest work at the #GRC.
Looking forward to the discussions, connections & science ahead!
🚨 Program Update: CADD GRS 2025!
We're excited to welcome @wpwalters.bsky.social as our keynote:
"AI in Drug Discovery – Revolution, Evolution, or Complete Nonsense?"
A perfect kickoff for an inspiring weekend in Portland!
🔗 www.grc.org/computer-aid...
#CADDGRS2025 #drugdesign #compchem
🏁 Program is live for the 2025 GRS on Computer-Aided Drug Design!
Overwhelming interest, top-tier abstracts, and a fantastic lineup incl. keynote Charlotte Deane + stellar mentorship panel. Can't wait!
🔗 www.grc.org/computer-aid...
#CADDGRS2025 #GRC #CompChem #DrugDesign
🚀 Exciting news! We're organizing a @cecamevents.bsky.social Flagship Workshop on Quantum Chemistry in Drug Design in Prague, Sept 8–10, 2025!
Join top experts from academia & industry. Few spots left for contributed talks!
📢 Apply now: www.cecam.org/workshop-det...
#compchem #cadd #QM #CECAM
Are single-structure scoring functions truly inferior to MD-based free energy methods? We recalibrated Wang dataset to test various approaches:
🔹 Conventional SFs (R² = 0.26)
🔹 SQM2.20 (R² = 0.47)
🔹 MD-based methods (R² = 0.52)
Check our newest preprint: doi.org/10.26434/che...
#cadd #compchem
(5/5) This work wouldn’t be possible without all incredible coauthors from indust&acad.
Thank you Christian Kramer, @jchodera.bsky.social , Kelly Damm-Ganamet, @bmix.bsky.social, Judith Günther, Uta Lessel, Richard Lewis, David Mobley, Eva Nittinger, @mattschap.bsky.social & @wpwalters.bsky.social!
(4/5) How do we push the field forward?
We need long-term, systematic benchmarking efforts with:
🔹 High-quality, unbiased datasets
🔹 Cross-industry collaboration
🔹 Open scientific discussions
Only then can we improve pose & activity prediction reliability in #SBDD!
#compchem #benchmarking
(3/5) Current pose- and activity- prediction benchmarks fall short.
Many datasets are biased, lack experimental validation, or fail to challenge methods properly. Blinded benchmarks ensure real-world performance, avoiding overfitting to known data.
#Benchmarking #compchem #cadd
(2/5) What makes an effective benchmark?
✅ Diverse datasets: drug-like molecules, hard targets, challenging interactions
✅ High-quality structures: clear density, reliable poses
✅ Robust affinity data: consistent Kd/Ki values
✅ Blinded validation: unbiased assessment
#compchem #benchmarking #cadd
🧵👇(1/5) Why does SBDD need better benchmarking?
Binding pose & activity predictions remain inconsistent, limiting drug discovery progress. A key issue? No standardized, widely accepted benchmark like CASP for protein structure prediction.
#Benchmarking #CADD #compchem
QM-based scoring function SQM2.20 - a universal yet computationally efficient physics-based method yielding reliable DFT quality affinity predictions in minutes!
www.nature.com/articles/s41...
+ new high-quality benchmark PL-REX dataset for the rigorous validation of your method
#compchem #cadd
New Perspective on Community Benchmarking in Structure-Based Drug Design (SBDD)!
#SBDD predictions need reliable benchmarks - diverse targets, high-quality affinity & structural data, and blinded validation. Let’s make it happen!
🔗 Read more: doi.org/10.1021/acs....
#DrugDiscovery #CompChem
