https://bcucco.gitlab.io/tags/posts/Recent content in Posts on AI-for-Science Systems (AIS²) LabHugoen-usSun, 08 Mar 2026 00:03:00 -0800https://bcucco.gitlab.io/post/molxprot/Tue, 21 Oct 2025 00:00:00 +0000https://bcucco.gitlab.io/post/molxprot/<p>What if you could predict how a drug molecules binds to its target protein, while being fast, reliable, and with more chemical interpretability than ever? That’s what <strong>MolXProt</strong> was designed to achieve.</p> <p>In the world of drug discovery, predicting the binding affinities between protein and ligand pairs is key. However, traditional methods such as docking and empirical scoring often fail to generalize to novel protein families and new chemistries.</p> <p>MolXProt was created to give a fresh new look to the problem: A <strong>transformer-based graph neural network</strong> that fuses graph-based ligand representations with protein language model embeddings via a <em>bidirectional multi-head cross-attention</em>. The results? We were able to obtain a scalable and interpretable model that delivers high-quality predictions and pave the way to understand how drugs and proteins “talk” at the atomic level.</p>https://bcucco.gitlab.io/post/matcssi/Mon, 04 Aug 2025 00:00:00 +0000https://bcucco.gitlab.io/post/matcssi/<p>For those working in computational chemistry/physics, it is clear that nowadays we have an immense zoo of quantum softwares and approaches, and these are constantly growing everyday. Moreover, several of these softwares can interact via often complex APIs, or requiring several pre- and post-processing steps. What if we could perform complex many-body quantum simulations workflows, from quasiparticle excitations to superconductivity, via a simple, unified, and reproducible architecture? That’s what <strong>MATCSSI</strong> (<a href="https://matcssi.tacc.utexas.edu/">https://matcssi.tacc.utexas.edu/</a>) is.</p>