New AI system designed to assist scientists in cancer-related discoveries

Research in the mountains of published cancer research could be facilitated for scientists through a new AI system.

The system, called LION LBD and developed by computer scientists and cancer researchers from the University of Cambridge, was designed to help scientists in search of cancer-related discoveries. This is the first literature-based discovery system to support cancer research. The results are reported in the newspaper bioinformatics.

Global cancer research is drawing huge amounts of money around the world, and the scientific literature is so huge that researchers are struggling to keep up: critical evidence, generating hypotheses, is often discovered long after their publication.

Cancer is a complex class of poorly understood diseases and the second leading cause of death in the world. The development of cancer involves changes in many molecules, reactions and chemical and biochemical pathways. In addition, cancer research is conducted in a wide variety of scientific fields, the descriptions of which vary across concepts.

"As a cancer researcher, even if you knew what you were looking for, thousands of papers are coming out every day," said Prof. Anna Korhonen, co-director of Cambridge Language Technology Lab, who led the development of LION LION in collaboration with Dr. Masashi Narita of the Cambridge Institute of Cancer Research UK and Professor Ulla Stenius of the Karolinska Institutet in Sweden. "LION LBD uses AI to help scientists keep abreast of new discoveries in their field, but could also help scientists make new discoveries by combining what we already know in the literature. linking sources that may seem unrelated. "

"LBD" in LION LBD stands for "Literature-Based Discovery", a concept developed in the 1980s that aims to make new discoveries by combining information from disconnected sources. The main idea of ​​the original version of LBD is that concepts that are never explicitly linked in the literature can be indirectly linked through intermediate concepts.

The LION LION system design enables real-time search to uncover indirect associations between entities in a database of tens of millions of publications while preserving the ability of users to explore each mention in its original context.

"For example, you may know that an anti-cancer drug affects the behavior of a specific pathway, but with LION LBD, you may find that a drug developed for a totally different disease affects the same way, "said Korhonen.

LION LBD is the first system developed specifically for the needs of cancer research. He focuses particularly on the molecular biology of cancer and uses advanced techniques of machine learning and natural language processing to detect references to the characteristics of cancer in the text. System evaluations have demonstrated its ability to identify undiscovered links and prioritize relevant concepts among potential connections.

The system relies on open data, open standards and open standards, and is available as an interactive web interface or programmable API.

Researchers are currently working on extending the scope of LION-LBD to include other concepts and relationships. They also work closely with cancer researchers to help and improve end-user technology.