Artificial Intelligence and Machine Learning in Chemistry and Molecular Biology
Several research groups at the Department of Chemistry and Molecular biology are active in the field of artificial intelligence (AI) and machine learning (ML). The work is related to a range of areas, from the development of novel drug candidates to the description of protein structures or interaction between proteins and carbohydrates.
Artificial intelligence (AI) and machine learning (ML) has had a tremendous impact in society during the last few years, and its use in natural sciences is no exception. Within chemistry and molecular biology, AI and ML approaches are for example used to develop new drug molecules, analyse data from large databases or predict the effect of different mutations on cellular function.
At the Department of Chemistry and Molecular Biology, several research groups develop and/or use different AI/ML based tools. This has led to collaborations with SciLifeLab and various academic groups and companies in the pharma area, as well as to patents and formation of spin-out companies.
Professor Leif Eriksson
Professor Leif Eriksson’s research group has developed the software tools i-TripleD and iPIN. Both tools are based on advanced AI. Using i-TripleD, researchers can now with extremely high accuracy identify novel drug molecules in a few hours. Earlier, the corresponding process took several months of supercomputer usage.
Dr Daniel Bojar
Dr Daniel Bojar’s research group uses AI to predict and understand glycan properties and functions. By developing the model LectinOracle, their work for instance facilitates protein-carbohydrate binding predictions at scale and in seconds, which otherwise may take months of experimental work.
Professor Gergely Katona
Professor Gergely Katona’s research group uses e.g. Bayesian methods to improve the interpretation of experimental data from for instance time resolved crystallography and to establish relationships between tests and results. The group has published several research articles where different AI-based methodologies are applied.
Professor Margit Mahlapuu
Professor Margit Mahlapuu collaborates with SciLifeLab in a project where data from experiments using DNA-encoded compound libraries will be used to train an AI model to identify new active compounds for the treatment of the liver disease NASH and liver cancer.
Professor Vladislav Orekhov
At the Swedish NMR Centre, Professor Vladislav Orekhov uses different AI-based methods such as neural networks to improve interpretation of NMR spectra and determine protein structures.
