Tools to leverage complex patient data for precision medicine
Instead of a few lab tests, thousands or even millions of molecular features can be measured simultaneously from a single patient with today’s technologies. With the right tools, this mass of data could reveal disease risk years in advance, predict disease progression and guide the best treatment choice for each patient.
Such tools for precision medicine are the research focus of Vittorio Fortino, who was recently appointed Professor of Bioinformatics and Machine Learning at the University of Eastern Finland’s Institute of Biomedicine. He describes his research as lying at the intersection of artificial intelligence, computational biology and biomedicine.
Omics technologies such as genomics, proteomics and metabolomics, have made it possible to analyse all of an individual’s genes, proteins, metabolites or other molecular features from a single sample. This generates enormous amounts of data for biomedical research, further multiplied in large patient cohorts and biobanks.
“My research group develops machine learning, data mining and other computational methods that help scientists and clinicians interpret this complex information.”
Machine learning-based biomarker panels
The discovery of new disease biomarkers is crucial for earlier and more precise diagnostics and even individually tailored treatments. “Traditionally, biomarkers are based on a single biological signal, such as the presence or level of a specific gene or protein. However, complex diseases cannot be explained by a single factor alone. Our AI models analyse large amounts of biological data to identify meaningful combinations of molecular signals, while ensuring the results are robust and clinically interpretable.”
These methods are applied in several national and international projects, spanning early disease detection and prognosis to treatment response and drug safety.
In a recent project, the Fortino group developed an AI-based computational toolbox, named BIODAI, to streamline the development and validation of machine learning-based biomarker panels.
“Clinicians are used to interpreting biomarkers by directly comparing measurements to reference values. Machine learning-based models must provide the results in a similarly reliable, interpretable and usable form as traditional diagnostic tools. They must also be constantly updated with new patient data to maintain their accuracy, something BIODAI is designed to ensure.”
The Research Council of Finland funded the project as well as a new proof of concept project where they are now testing the BIODAI toolbox in real-world settings.
“We apply these approaches to cancer patient data from Kuopio University Hospital to predict response to immunotherapy. At the industrial level, we are using BIODAI to develop machine learning-based biomarkers that can identify patients at risk of more severe disease trajectories, helping ensure timely and appropriate treatment.”
From drug repurposing to environmental risk assessment
Fortino’s group is involved in a variety of precision medicine-related projects targeting different diseases and multiple aspects of drug discovery, ranging from drug repurposing and sensitivity prediction to in silico toxicity assessment and the discovery of new putative drug targets.
“One important approach is pharmacogenomics, where we try to stratify cancer patients in subgroups based both on their molecular make-up and their sensitivity to treatments. To do this, we use experimental cancer models combined with large-scale drug screening data to learn patterns of drug sensitivity and transfer this knowledge to patients. This work is supported by Jane and Aatos Erkko Foundation.”
This may enable researchers to better understand why some cancer patients respond to treatments while others do not.
“We can also identify existing medicines that could be repurposed to fight new types of cancer. For patients with limited drug response, new drug combinations could be designed to improve outcomes.”
Fortino’s group also works in the field of toxicogenomics, dissecting the molecular mechanisms underlying diseases caused by exposure to harmful compounds.
“In the EU-funded EDCMET and NEMESIS projects, our expertise is used to elucidate the mechanisms linking exposure to endocrine disrupting chemicals to adverse metabolic effects. This work contributes to the development of future in silico methods to assess the harmful potential of untested compounds while reducing the need for animal testing.”
Multidisciplinary collaboration and education
Fortino points out that as a bioinformatician with a background in computational science, his work relies heavily on interdisciplinary collaboration with researchers from life sciences and clinical medicine.
“The challenge is to develop machine learning solutions for biomedical problems while avoiding false findings. In the era of deep learning and generative AI, new models are emerging rapidly, but their real-world reliability is not always carefully assessed.”
“There is also a growing need for more multidisciplinary educational pathways that combine biology, medicine, statistics, computer science and artificial intelligence. Future bioinformaticians should develop both the biological understanding and the computational skills required to work with complex biomedical data,” Fortino says.
“At the same time, the rapid growth of AI tools requires careful guidance in education. While AI can assist with tasks such as coding, students must still develop a strong understanding of the theoretical foundations of data analysis and methods they use.”
Vittorio Fortino
Professor of Bioinformatics and Machine Learning, University of Eastern Finland, 1.1.2026–
- Master's Degree in Computer Science, University of Salerno, Italy, 2009
- PhD in Bioinformatics and Systems Biology, University of Salerno, Italy, 2013
Key roles:
- Associate Professor (tenure track) of Health Bioinformatics and Big Data Analytics, University of Eastern Finland, 2021–2025
- Assistant Professor (tenure track) of Health Bioinformatics and Big Data Analytics, University of Eastern Finland, 2018–2021
- Academy Research Fellow, University of Eastern Finland, 2020–2024
- Postdoctoral Researcher, University of Tampere, 2017–2018
- Postdoctoral Researcher, University of Helsinki, 2016–2017
- Postdoctoral Researcher, Finnish Institute of Occupational Health, 2013–2016
- Data Scientist, University of Salerno, Italy, 2012–2013
