Sign in

Leonardo Medrano

@lmedranos88.bsky.social
168 followers 445 following 15 posts

Computational physicist/chemist at @casusscience.bsky.social, Germany | Chemical Physics, Machine learning, NanoPhononics, Computational modeling, Materials Science 🇵🇪

PostsRepliesMedia
Leonardo Medrano @lmedranos88.bsky.social · 29/09/2026
🚨Preprint alert: We introduce our #QALPA framework (Quantum-Aware Learning for Property-space Augmentation)! 🎉 QALPA combines conditioned equivariant diffusion modeling with multiple electronic-structure methods to augment QM datasets along a trajectory in a property space.
https://arxiv.org/abs/2609.16527
100
Leonardo Medrano @lmedranos88.bsky.social · 25/09/2026
🎉 If you have ever wondered how much CO2 we emit when developing a quantum-mechanical dataset or training an AI model, we invite you to read our recent Perspective on “Sustainable Machine Learning” recently published in @digital-discovery.rsc.org! 👏 Many thanks to everyone for their contributions!
010
Reposted by Leonardo Medrano
Digital Discovery @digital-discovery.rsc.org · 06/02/2026
☀️ Don't miss this #DigitalDiscovery read! ✨ Leonardo Sandonas et al. introduce QUED, a hybrid QM/ML framework that integrates molecular structure and electronic information to deliver accurate predictions of physicochemical and biological properties. ➡️
doi.org
Assessing the performance of quantum-mechanical descriptors in physicochemical and biological property prediction
Machine learning (ML) approaches have drastically advanced the exploration of structure–property and property–property relationships in computer-aided drug discovery. A central challenge in this field is the identification of molecular descriptors that can effectively capture both geometric- and electronic structur
021
Reposted by Leonardo Medrano
LuxProvide @luxprovide.bsky.social · 14/08/2025
New on #HPCSummerQuest: EquiDTB blends quantum chemistry with equivariant AI to reach DFT-level accuracy for large, flexible molecules. Trained on #MeluXina GPUs, cutting runtimes from weeks to days. ⚡️🧬🧠 Read more 👉 www.luxprovide.lu/advancing-de...
121
Leonardo Medrano @lmedranos88.bsky.social · 15/07/2025
🗣️The #ieeenanoperu Chapter, in collaboration with #IEEE Nanotechnology Council, is organizing the first IEEE #LatinAmerican Conference on #Nanotechnology ( #ieeelanano), to be held in the historic city of Cusco, #Peru, from November 4-7, 2025. 👉 ieee-lanano.org @ieeexplore.bsky.social #IYQ2025
010
Reposted by Leonardo Medrano
Microsoft Research @msftresearch.bsky.social · 10/07/2025
Today in the journal Science: BioEmu from Microsoft Research AI for Science. This generative deep learning method emulates protein equilibrium ensembles – key for understanding protein function at scale. www.science.org/doi/10.1126/...
110849
Leonardo Medrano @lmedranos88.bsky.social · 07/07/2025
🚨Our short review “Recent Advances in #MachineLearning and #CoarseGrained Potentials for #BiomolecularSimulations” has been accepted in @biophysj.bsky.social @cellpress.bsky.social! Many thanks to all authors for their contributions! This was a fantastic collaboration!👏 www.cell.com/biophysj/ful...
110
Leonardo Medrano @lmedranos88.bsky.social · 02/06/2025
👋Just two weeks left to register for the #SusML Workshop 2025 in Dresden! susml.net Have a look at our exceptional lineup of speakers that will discuss current ML methods for the sustainable exploration of chemical spaces. #Psik @tudresden.bsky.social @digital-discovery.rsc.org @mpipks.bsky.social
041
Leonardo Medrano @lmedranos88.bsky.social · 14/05/2025
From foundational datasets like QM9, QM7-X, ANI, Aquamarine, and GEOM (among others) to the recently published #QCML and now #TheOpenMolecules2025! The exploration of the #ChemicalSpace through #QuantumMechanical properties has progressed remarkably over the past five years. 😀 #sustainableML
000
Reposted by Leonardo Medrano
Volker Blum @aimsduke.bsky.social · 02/05/2025
The biggest paper I was ever part of appeared on arXiV today: "Roadmap on Advancements of the FHI-aims Software Package". Over 20 years of work. Immensely grateful to the 200+ people on this paper, who pushed our ability to simulate materials forward! #chemsky #compchemsky arxiv.org/abs/2505.00125
arxiv.org
Roadmap on Advancements of the FHI-aims Software Package
Electronic-structure theory is the foundation of the description of materials including multiscale modeling of their properties and functions. Obviously, without sufficient accuracy at the base, relia...
42810
Leonardo Medrano @lmedranos88.bsky.social · 25/04/2025
🚨Check out our recent #preprint on advancing Density Functional Tight-Binding method with equivariant NNs. We have been developing this project for a while, and we now present the results that highlight the enhanced scalability/transferability of our DFTB+ML approach. 🌐 chemrxiv.org/engage/chemr...
010
Reposted by Leonardo Medrano
MDDB @mddbeu.bsky.social · 04/04/2025
📢 Our article calling for a #FAIR database for #MolecularDynamics simulation data has now been peer-reviewed and published in @naturemethods.bsky.social 📖 Read it here: rdcu.be/ef6YX 📝 Support the statement: bit.ly/3zVS3qm #MDDB #FAIRdata #collaboration
03821
Leonardo Medrano @lmedranos88.bsky.social · 15/03/2025
👋Hallo there! The registration for the SusML workshop is OPEN! Join us in discussions on topics such as data-efficient ML-based methodologies and the inverse property-to-structure problem. See you at @tudresden.bsky.social in Germany! 👉More information: susml.net Stay tuned for more updates!😉
1135
Leonardo Medrano @lmedranos88.bsky.social · 01/03/2025
👋The MORE-Q dataset is finally out in Scientific Data! 😃 We have performed extensive electronic structure calculations to generate #quantumechanical property data for building blocks of mucin-derived olfactory #sensingdevices. 🌐You can read more about MORE-Q at: www.nature.com/articles/s41...
nature.com
MORE-Q, a dataset for molecular olfactorial receptor engineering by quantum mechanics - Scientific Data
Scientific Data - MORE-Q, a dataset for molecular olfactorial receptor engineering by quantum mechanics
010
Leonardo Medrano @lmedranos88.bsky.social · 08/02/2025
👋Preprint: check out our contribution to this short review where we discuss recent efforts toward developing robust #machinelearning potentials for #biomolecularsimulations with #quantummechanical accuracy. 👀 👉The preprint is available on @chemrxiv.bsky.social: chemrxiv.org/engage/chemr...
010
Reposted by Leonardo Medrano
profvlilienfeld.bsky.social @profvlilienfeld.bsky.social · 03/02/2025
Pumped about #MachineLearning the adaptive exact exchange admixture in hybrid #DFT approximations: It can even cure the infamous spin-gap problem (see below). Just out in @ScienceAdvances with D Khan, A Price, B Huang and M Ach! @uoft.bsky.social #CompChem www.science.org/doi/10.1126/...
0135
Leonardo Medrano @lmedranos88.bsky.social · 19/01/2025
🚨The first preprint of the year is out on @chemrxiv.bsky.social! Great collaboration with Mirela, Alexander, Peter et al.!!👍 chemrxiv.org/engage/chemr... We introduce the QUID benchmark framework for large non-covalent systems, capturing frequent ligand-pocket interaction types.
020
Reposted by Leonardo Medrano
Microsoft Research @msftresearch.bsky.social · 16/01/2025
Microsoft researchers introduce MatterGen, a model that can discover new materials tailored to specific needs—like efficient solar cells or CO2 recycling—advancing progress beyond trial-and-error experiments. www.microsoft.com/en-us/resear...
16226