Teaching machines to read life:BioGeMT explores AI in Bioinformatics

The University of Malta's Valletta Campus hosted the Summer School "AI/ML Applications in Bioinformatics 2026" from 22 to 25 September 2026. Organised by the EU-funded Bioinformatics for Genomics in Malta (BioGeMT) project, the four-day programme was a resounding success, attracting over 20 participants from universities across Europe.

Artificial intelligence is rapidly reshaping how biological research is done. Models that can predict the three-dimensional structure of a protein, read the regulatory grammar of the genome, or design entirely new molecules are no longer niche tools but increasingly part of everyday practice in the life sciences.

Keeping pace with this shift requires researchers who understand not only how to use these methods, but how they work, where they fail, and how to interpret their outputs. The Summer School was designed with exactly this in mind, combining theoretical foundations with hands-on practical workshops that took participants from core concepts to the current state of the art.

The program opened with a plenary talk by Dr Panagiotis Alexiou, followed by a lecture and an afternoon workshop on AI fundamentals led by Vlastimil Martinek, giving all participants a common grounding in the core methods of machine learning.

The second day turned to foundation models, the large pre-trained models that are transforming genomics in much the same way that large language models have transformed text. Dr Alexiou introduced foundation models and self-supervised learning in biology, before Andrea Gariboldi and Dr Alexiou led a hands-on workshop on AlphaGenome and related models. In the afternoon, Carlos Gafa addressed eXplainable Artificial Intelligence (XAI), a critical topic as researchers and clinicians increasingly need to understand why a model reaches a given prediction before they can trust it. The day closed with a conference dinner, giving participants, speakers and collaborators the chance to network and continue the discussion.

The third day was dedicated entirely to agentic systems in bioinformatics, one of the fastest-moving areas in the field. In an extended, full-day workshop, Dimosthenis Tzimotoudis and Andrea Gariboldi of the Agentomics team showed how AI agents can plan and carry out multi-step bioinformatics analyses, and explored both the promise and the limitations of this emerging approach.

On the final day, Prof. Jean-Paul Ebejer delivered a two-part session on AI for protein structure, function and design, tracing the developments inspired by AlphaFold and beyond, from structure prediction to the generative design of new proteins. He also led a session on computer-aided drug design, showing how recent advances in AI and other computational methods can be used to identify and optimise new therapeutic molecules.

The Summer School concluded with the BioGeMT Symposium on the afternoon of the last day, where collaborators and partners from across the project presented their research and discussed future directions for bioinformatics and genomics in Malta and beyond.

The organisers would like to thank all the speakers, from the University of Malta, the Agentomics team, and international collaborators, for their generous contributions, and all the participants for their enthusiasm and engagement throughout the week. Thanks also go to the University of Malta's Conferences and Events Unit for their support in organising the event.

The Summer School forms part of BioGeMT's wider mission to strengthen bioinformatics and genomics research capacity in Malta, to train the next generation of researchers in data-driven life science, and to build lasting links with leading research institutions across Europe.

About BioGeMT

Bioinformatics for Genomics in Malta (BioGeMT) is an ERA Chair project funded by the European Union under Horizon Europe's Widening Participation and Spreading Excellence programme (grant agreement no. 101086768), with an EU contribution of almost €2.5 million. Coordinated by the University of Malta and running from December 2022 to November 2027, the project hosts an expert bioinformatician and team within the Centre for Molecular Medicine and Biobanking, working alongside the University's molecular genetics research groups.

Its aim is to build lasting bioinformatics and machine learning expertise at the University of Malta, so that patient-derived multi-omic and clinical data can be turned into a better understanding of disease.

Malta's population genetics and large family structures make it well placed to shed light on the genetic basis of disease, and the project builds on existing biobanks and sample collections, including one with more than 2,000 whole genomes and RNA-Seq datasets. In the longer term, BioGeMT aims to establish a specialised bioinformatics centre in Southern Europe, strengthen Malta's competitiveness for international research funding, and support the national strategy to grow ICT for health.