Self-learning machines will show the way to better drug therapies

“AI research can quite possibly already have the tools we need,” says Per Larsson, researcher at Uppsala University that, with the help of machine learning, aim to develop methods to calculate the solubility and absorption of oral peptide drugs into blood circulation.

(Image removed) Per Larsson, researcher at SweDeliver and Uppsala University

Self-propelled vacuum cleaners and auto-generated Spotify playlists, artificial intelligence, AI, is already well integrated into our everyday lives. Now the self-learning machines are entering drug development. A robotised laboratory was recently inaugurated at Uppsala University's Faculty of Pharmacy, and in January 2022, senior lecturer Per Larsson will launch a project aiming to, via machine learning, enable predictions on how orally administered peptide drugs will act in the colon.

“Using physical principles, we can already calculate relatively accurate how smaller molecules will behave in an uncomplicated aquatic environment. However, we still face challenges in predicting how macromolecules might react in complex intestinal environments. Now the rapid development in AI is opening new doors in a number of research fields, and here we have realistic opportunities to take drug delivery research to the next level.”

From January and twelve months onwards, Per Larsson will develop his ideas at AI4Research, a five-year initiative with the aim to establish Uppsala University at the frontline of research and education on our digital future. The AI4Research environment is an interdisciplinary melting pot where researchers from all disciplinary domains of the university meet to generate new AI-related perspectives and projects that will hopefully take on a life of their own at their respective home departments.

(Image removed) AI4Research was inaugurated 2 October 2020

“We are seven researchers who have awarded grants for 2022, and in addition to the availability of expert support to develop my project, it is extremely inspiring to be able to lift my gaze and see how AI can be applied in other areas of research. The position at AI4Research also includes the opportunity to invite my PhD students, so hopefully this is also an opportunity to integrate AI into our research environment in the long term.”

The meeting between mathematical and medical skills already creates great synergy effects. In Uppsala, algorithms are being developed with the capacity to extract completely new knowledge from the Swedish National Quality Registers. In healthcare, there are computer systems that interpret ECG results and, using machine learning, learn to solve tasks for which they are not explicitly designed. In short, we live in a time where science is rapidly bringing us towards the realization of several intergenerational utopias.

“Oral drug delivery faces major challenges. Nevertheless, it is entirely conceivable that AI research already has the tools needed to answer our questions: How will macromolecules interact with other formulations? Can they get from the colon via the cell membrane and enter blood circulation, and how do we achieve sufficient solubility and absorption? Within a year, I hope that we have taken several important steps forward.


  • is tasked with strengthening, renewing and developing research in AI and machine learning.
  • offers researchers funding for a sabbatical period of at least half time for a maximum of one year.
  • is a resource for the entire University, and researchers in all disciplinary research domains at Uppsala University will be included throughout the project period.



(Image removed) Per Larsson, Senior lecturer
Department of Pharmacy

text: Magnus Alsne, photo: Mikael Wallerstedt, AI4Research