Dr. Tomasz Kościółek wins prestigious FNP FIRST TEAM grant to develop the world’s first generative foundation model of the human gut microbiome.

Dr. Tomasz Kościółek wins prestigious FNP FIRST TEAM grant to develop the world’s first generative foundation model of the human gut microbiome.

Sano – Centre for Computational Medicine proudly announces that Dr. Tomasz Kościółek has been awarded the First Team grant of the Foundation for Polish Science (FNP).

Sano – Centre for Computational Medicine proudly announces that Dr. Tomasz Kościółek has been awarded the First Team grant of the Foundation for Polish Science (FNP). The funding will support a project titled: ‘A generative foundation model of the human gut microbiome for digital twin development and biomedical applications’ (original pol. “Generatywny model fundamentalny ludzkiego mikrobiomu jelitowego do tworzenia cyfrowych bliźniaków i zastosowań biomedycznych”). 

This pioneering initiative aims to build the first generative foundation model of the human gut microbiome, enabling the creation of personalized digital twins for biomedical research, diagnostics, prevention, and therapy optimization. 

A breakthrough in computational biology and personalized medicine 

The project will develop a next-generation model — microbiomeGPT — trained on tens of thousands of metagenomic data points from diverse populations and disease states. Unlike current microbiome analysis tools, microbiomeGPT will: 

  • learn both taxonomic and functional characteristics of microbial communities, 
  • identify biomarkers for diseases such as inflammatory bowel disease, colorectal cancer, diabetes, obesit, Parkinson’s disease, and mental health disorders, 
  • simulate future states of the microbiome and predict the effects of diet, therapy, or interventions, 
  • generate digital twins of patients to support precision medicine and pharmaceutical R&D. 

This generative, multi-parametric approach has the potential to revolutionize the way clinicians, researchers, and industry understand and manipulate the microbiome. 

International excellence and strategic partnerships 

The project brings together leading research institutions and industry partners: 

  • Prof. Elin Org, University of Tartu (Estonia) – providing access to the Estonian Microbiome Cohort and expertise in population-scale microbiome studies. 
  • Prof. Ewa Szczurek, Helmholtz Munich & University of Warsaw – advising on deep generative AI models and multi-omics integration. 
  • Diagnostyka S.A., Poland’s largest medical laboratory network – supporting translational testing, validation, and pathways to future clinical application. 

Work on the project will be conducted at Sano’s Structural and Functional Genomics Research Group, beginning July 1, 2026 and continuing through December 31, 2029. 

Strengthening AI for health 

Dr. Tomasz Kościółek— is a recognized bioinformatician and an expert in microbiome research, and machine learning. His contributions include the development of DeepFRI, Qiita, and numerous high-impact publications across Nature, Nature Communications, Nature Biotechnology, and more. 

His First Team project positions Sano at the forefront of global efforts to build AI-driven biological models, aligning with European priorities in digital health, precision medicine, and computational simulation. 

K. Nicholson, Head of Innovation: 
“This project opens a completely new chapter in microbiome science. For the first time, we will be able to model the human gut ecosystem not just by observing it, but by generating new, hypothetical scenarios—much like large language models do with text. microbiomeGPT will enable clinicians and researchers to predict disease risk, simulate treatment outcomes, and design personalized interventions before they happen in the real world.” 

#FunduszeUE