Artificial intelligence helps in the screening and design of antitumoral drugs carried out at the CNIO's Experimental Therapeutics program. Credit: Laura M. Lombardía / CNIO
The research community recognizes the arrival of artificial intelligence (AI) as a revolution, a “paradigm shift” that affects all areas of research. In oncology it is already being used to better detect and diagnose cancer, and in the development of more effective treatments with fewer side effects. But this is just the beginning.
The Spanish National Cancer Research Center (CNIO) uses AI in areas such as genomic and big data analysis; image analysis; protein structure prediction; and anti-tumor drug discovery. And it will soon have a new artificial intelligence unit thanks to an allocation of 4.6 million NextGenerationEU European funds for the promotion of digital talent (managed by Red.es, of the Ministry for Digital Transformation and Public Administration).
For Maria A. Blasco, CNIO Scientific Director, “AI is essential to be able to diagnose and treat cancer more efficiently. The CNIO’s scientific priorities have always evolved in parallel with changes in scientific paradigms and new technologies. Artificial intelligence is one of these new paradigms”.
These are some applications of AI in CNIO programs:
- Development of new drugs. The CNIO’s Experimental Therapeutics program is conducting research to discover and develop new anti-tumor drugs. For its director, Joaquín Pastor, artificial intelligence “has the potential to revolutionize the drug discovery process, making it faster, cheaper and more efficient”. At the CNIO, AI helps in drug screening and design, the prediction of polypharmacology –drugs that act on different targets at the same time– and the reuse of drugs, among other applications.
- Identify new targets: “AI is allowing, for example, to design artificial proteins that have not existed before in nature, and that we can use in the development of new therapies, to interfere with specific cellular machines”. Óscar Llorca,director of the Structural Biology program at the CNIO.
- Finding the relevant information in the sea of data: “The amount and complexity of the biological data we handle [thousands of variables, tens or hundreds of samples], AI helps us in their analysis, from the detection of low-quality samples to the detection of patterns or the contextualization of the results”. Leonardo Daniel Garma, researcher at the CNIO Breast Cancer Clinical Research Unit. This group leads the Digital Twins Project, which seeks 300 women with advanced cancer to create virtual models of their disease, which can help future patients.
- Open avenues of research by associating different kinds of data. The most widely used technologies in cancer research, such as genomics and proteomics or imaging technologies, generate huge amounts of data. AI is essential to draw valid conclusions from them, “but it also helps to link data from very different technologies, which represents an additional dimension of data analysis, unprecedented in history”. Fernando Peláez, director of the CNIO Biotechnology program.
- Knowing how the tumor will evolve. In the Human Cancer Genetics program they use AI to predict tumor prognosis from the expression of lists of genes or molecular signatures, and from patterns analyzed in whole tumor sections, explains Mercedes Robledo, head of Endocrine Hereditary Cancer Group at the CNIO. Her group participates in the IMPact-VUSCan project, which uses AI to classify genetic variants that could be related to the origin of tumors