BEGIN:VCALENDAR VERSION:2.0 PRODID:researchseminars.org CALSCALE:GREGORIAN X-WR-CALNAME:researchseminars.org BEGIN:VEVENT SUMMARY:Alvaro Gargantilla Becerra (National Centre for Biotechnology (CNB )) DTSTART:20260914T140000Z DTEND:20260914T143000Z DTSTAMP:20260605T005623Z UID:MicroBiotechDBTL/11 DESCRIPTION:Title: ML4SD: An Active Learning Framework for Boosting DBTL Cycles in Strain Design\nby Alvaro Gargantilla Becerra (National Centre for Biot echnology (CNB)) as part of Seminar on Microbial Biotechnology: Developing the Conceptual Framework of the DBTL Cycle\n\n\nAbstract\nModern biomanuf acturing relies on iterative Design-Build-Test-Learn (DBTL) cycles\, yet t he "Learn" phase is frequently the weakest link due to the limited predict ive power for complex biological systems. Conventional approaches are unab le to cope with the "combinatorial explosion" of the genetic design space\ , particularly when engineering growth-coupled phenotypes through gene kno ckouts.\n\nIn this presentation\, we propose ML4SD\, a computational frame work that integrates machine learning into the DBTL cycle with the objecti ve of accelerating strain design. The present approach utilizes gcSwarms\, a novel binary particle swarm optimisation algorithm that generates exten sive and diverse design libraries with a view to enhancing machine learnin g model generalization. The employment of an active learning strategy that balances exploration and exploitation enables ML4SD to iteratively refine knockout recommendations whilst identifying key metabolic interventions t hrough the utilization of Shapley-based explainable AI.\n\nA partial valid ation of ML4SD was made using Nylon-6 precursor production in Pseudomonas putida\, resulting in a yield that was five times higher than that obtaine d with the wild type. In addition\, ML4SD has been shown to exhibit superi or data- and resource-efficiency in comparison to its data source generati on. Overall\, this methodology establishes a robust and user-friendly plat form for the autonomous development of high-performing microbial cell fact ories.\n\nÁlvaro Gargantilla Becerra is a biochemist who has transitioned into the roles of bioinformatician and data scientist\, with a specialisa tion in systems biology. Recently\, he was awarded his PhD in Molecular B iosciences from the Universidad Autónoma de Madrid. His academic journey commenced with a degree in Biochemistry\, followed by a research stay in t he domain of synthetic biology at Cardiff University. During his Master's programme in Industrial Biotechnology\, he developed computational screeni ng tools using genetic algorithms and neural networks. The primary focus o f his doctoral research at CNB-CSIC was on the utilization of bacterial he terogeneity for the purpose of bioprocess optimisation through the impleme ntation of machine learning (ML)-boosted DBTL cycles. This research result ed in the development of the ML4SD framework and the gcSwarms algorithm\, which have the potential to accelerate sustainable biomanufacturing throug h data-efficient strain design.\n LOCATION:https://researchseminars.org/talk/MicroBiotechDBTL/11/ END:VEVENT END:VCALENDAR