1/10 Phylogenetic trees are computationally expensive, combinatorially brittle, and—for molecular clock dating—rather unnecessary. Introducing ChronAeon: tree-free molecular clock dating in seconds, powered by distance geometry and our HyphAeon transformer foundation model. 🧵

💥 2026 A2B2C CONFERENCE #XVCAB2C 📣 LAST 9 DAYs TO SEND YOUR ABSTRACT 📌 November 9-12, 2026 👉More info: 2026.a2b2c.org.ar

BADACO 2026: Primeras Jornadas de construcción, gestión y diseño de BAses de DAtos COlaborativas, accesibles y abiertas 22 y 23 de octubre, presencial y gratuito. Resúmenes hasta el 15/09 jornadasbadaco.org/

💥 2026 A2B2C CONFERENCE #XVCAB2C 📣 ABSTRACT SUBMISSION IS OPEN 📌 November 9-12, 2026 👉More info: 2026.a2b2c.org.ar

1/ Curious how protein language models such as @biohub's ESM-C represent structural information from sequence alone? Come explore this interactive demo we built with @GoodfireAI's new Silico research platform and see for yourself!

1/ What could we discover if we had a binder for every protein in the human proteome? We are excited to introduce the Bindome, an open and free resource of over 300,000 designed protein binder candidates against over 8,000 human target proteins. 🧵

La @UNESCO lanzó su Plataforma de #CienciaAbierta: un recurso libre para hacer visible, reutilizable y medible el conocimiento científico abierto. Incluye buscador de publicaciones y dashboard con datos de 81 países. 👉 open-science.unesco.org/

#BADACO2026 1ras Jornadas de construcción, gestión y diseño de bases de datos colaborativas, accesibles y abiertas. 📆 22 y 23 de octubre de 2026 📍 Universidad Nacional de Quilmes 🇦🇷 🎟 Inscripción gratuita 📝 Envío de trabajos hasta el 15/09 🔗 jornadasbadaco.org

📢Se encuentra abierta la convocatoria 2026 Programa FRIDA para postular proyectos vinculados al desarrollo de Internet programafrida.net/ Conoce las líneas de financiamiento y ejes temáticos. ¡Anímate a postularte!

🌎 Grateful for the inspiring #LABIMeeting2025! 🍎 We joined as a Partner Organization and shared the workshop 🎙️ “Unlocking Funding Opportunities for Scientists – Grant Writing” 💜 Thanks @LatamBioimaging community for the invitation and exchange! doi.org/10.5281/zenodo.17546…

A blueprint for a fairer system! Our new article explores how the GlobalSouth is pioneering equitable, multilingual, and community-driven models of scholarly publishing. #OpenScience #SciComm Read it here, via @eLife @eLifeCommunity 📖 elifesciences.org/articles/1…

📣 Estamos felices de anunciar que se acerca la IJAB2C Este es un evento abierto a toda la comunidad y nos gustaría invitarte a participar 🤗 🗓️ Envío de Resumenes hasta el 10 de Octubre ➕ Información: bit.ly/IJAB2C

🌎 Ya está disponible nuestro preprint sobre acceso al financiamiento en América Latina, desarrollado gracias a los aportes de nuestra comunidad 🙌 📄 Texto completo (en inglés): doi.org/10.31222/osf.io/68ws… 💙 Gracias a quienes participaron y apoyan este proyecto.

📝VI Jornadas de Investigadores e investigadoras en Formación en Ciencia y Tecnología 📣Se realizará en la UNQ los días 13 y 14 de noviembre. 🔗+ INFO: bit.ly/4l4dveP

🚀🔍 En la #DiTella buscamos investigadores/as con estudios de posgrado y perfiles especializados en producción, operaciones, Industria 4.0, inteligencia artificial, mecatrónica, robótica, control, eficiencia energética, sustentabilidad y más. 👉 Info en bit.ly/4dKHXbJ.

🙋‍♀️🙋‍♂️ ¿Quieres compartir tu opinión sobre el acceso a financiamiento para investigación? 🌎 Estamos diseñando una formación que acerque herramientas y experiencias concretas para facilitar el acceso a fondos 📝 Completa la encuesta y ayúdanos a compartirla: es.surveymonkey.com/r/financ…

Pregunta de historia del modelo cientifico: El acuerdo Bretton Woods puso como 🪙 de referencia internacional al 💵. Esto se dio porque 🇺🇸 habia sido ganador de la WW II. Pero... cuando fue el "Bretton Woods" donde se decidio que los papers se escriben en ingles, y por que?

We investigated how LTSV40 inhibits E2F binding to a key regulator of the cell cycle (retinoblastoma protein). @NPalopoli @Carlalpf91 @jpalma1503 proliferationhttps://www.frontiersin.org/journals/chemical-biology/articles/10.3389/fchbi.2025.1538350/full

Assessing scoring metrics for AlphaFold2 and AlphaFold3 protein complex predictions 1. This study benchmarks commonly used scoring metrics to assess the quality of protein complex models predicted by AlphaFold2 (via ColabFold) and AlphaFold3, revealing key differences in their ability to evaluate heterodimeric interfaces accurately. 2. Using 325 high-resolution heterodimeric structures, the authors found that ColabFold with templates and AlphaFold3 produced similarly high proportions of correct models, both outperforming ColabFold without templates, especially in generating high DockQ-scoring models. 3. Interface-specific scores—particularly ipTM and model confidence—emerged as the most reliable metrics for evaluating protein complex models, consistently showing the highest correlation with DockQ across all datasets. 4. Global scores like pLDDT and PAE were less effective than interface-focused metrics such as ipLDDT and iPAE. Notably, pDockQ2 had poor correlation with actual model quality, often misclassifying high-quality models. 5. To address these inconsistencies, the authors developed C2Qscore, a combined score using linear regression over multiple interface-specific metrics. C2Qscore outperformed individual metrics and unweighted averages, especially in X-ray-validated datasets. 6. The study derived method-specific cutoffs for key metrics like ipTM to better classify models as correct or incorrect, including within the “gray zone” (ipTM = 0.6–0.8) defined by AlphaFold. These thresholds improved assessment precision and reduced false positives. 7. While AF3 produced more consistent results across replicates, CF-T (ColabFold with templates) generated a slightly higher number of top-quality predictions. However, AF3 models were more reliable when evaluated across a broader quality range. 8. In complex cryoEM-derived assemblies, the study highlights limitations of all evaluated methods: when multiple dimeric configurations exist, DockQ often misclassifies good models due to reference mismatches, but C2Qscore remained relatively robust. 9. The authors integrated C2Qscore into the ChimeraX plug-in PICKLUSTER v2.0, enabling researchers to visualize and assess predicted interfaces with detailed, customizable metrics—including support for AlphaFold3 models. 10. This work provides a practical guide for interpreting and comparing protein complex predictions across AF2/ColabFold and AF3, offering scoring thresholds, benchmarking insights, and tools for improving confidence in computational structural biology. 💻Code: gitlab.com/topf-lab/picklust… 📜Paper: biorxiv.org/content/10.1101/… #AlphaFold2 #AlphaFold3 #proteincomplex #structureprediction #bioinformatics #modelassessment #interfaceprediction #ColabFold #scoringmetrics #AI4Science

PepPCBench is a Comprehensive Benchmark for Protein-Peptide Complex Structure Prediction with AlphaFold3 1. PepPCBench introduces the most comprehensive benchmark to date for evaluating protein-peptide complex structure prediction using AlphaFold3 (AF3) and other PFNNs, addressing a critical gap in peptide drug discovery tools. 2. At its core is PepPCSet, a rigorously curated dataset of 261 non-redundant protein-peptide complexes (5–30 residues), designed to be independent of any AlphaFold3 training data and structured to reflect real-world docking challenges. 3. The benchmark compares five PFNNs (AF3, AFM, RoseTTAFold-All-Atom, Chai-1, HelixFold3) using standardized metrics like DockQ, L_RMSD, F_nat, MolProbity score, and Rosetta binding energy, offering a multi-dimensional assessment. 4. AF3 consistently outperforms its peers in terms of docking accuracy, structure quality, and ranking power, but still struggles with longer or flexible peptides and real-world virtual screening tasks. 5. The study highlights the structural "memorization" issue in PFNNs, showing that AF3 performs significantly better on peptide complexes similar to its training data, with DockQ dropping from 0.70 (train set) to 0.46 (test set). 6. Peptide length strongly impacts prediction: for peptides under 20 residues, docking accuracy decreases with length due to conformational flexibility; interestingly, AF3 regains accuracy for peptides longer than 20, aligning them more with protein folding behavior. 7. AF3 performs best on “easy” docking cases but still shows solid results on “difficult” complexes, especially when peptides are longer and exhibit protein-like structural constraints. 8. Confidence-related metrics like ipTM and actifpTM correlate well with docking quality, especially for AF3, enabling better model ranking. Combining these with Rosetta binding energy and structural relaxation improves predictive discrimination. 9. Enhanced sampling (increasing seeds and model replicates) has minimal impact on average predictive performance but raises the theoretical success limit, indicating potential for better sampling-aware scoring strategies. 10. AF3’s structural outputs are geometrically and stereochemically robust, with MolProbity and Clash Scores approaching those of native PDB structures—highlighting its utility for downstream applications. 11. A case study on MHC-II peptide binders shows AF3’s structure-based rankings correlate well with experimental affinities. actifpTM proved most reliable for recovering high-affinity peptides, suggesting practical relevance for screening workflows. 12. Despite AF3’s impressive performance, challenges like scoring function optimization and training data bias remain. PepPCBench serves not only as a benchmark but as a practical guide for using PFNNs in peptide-based drug design. 13. The authors emphasize transparency and reproducibility, providing all data, processing protocols, evaluation scripts, and structural models to facilitate adoption and future development. 💻Code: github.com/zhaisilong/PepPCB… 📜Paper: biorxiv.org/content/10.1101/… #peptidedesign #AlphaFold3 #proteinpeptide #structuralbioinformatics #drugdiscovery #molecularmodeling #bioinformatics #PFNN #benchmarking