A team of bioengineering and computing experts outlines how the creation of text mining and Artificial Intelligence tools could advance biomaterials research and development

19 June 2020
The application of text mining technologies will increase the ability to extract information from the growing biomaterials literature, and deep learning tools will help to identify implicit links between substantiated information, as well as make predictions and recommendations.

This comment article by authors from the Barcelona Supercomputing Center, the Universitat Politècnica de Catalunya, and the Institut de Bioingeniería de Catalunya has been published today in Nature Review Materials magazine.

Nature Review materials magazine has published today an article signed by scientists from the Barcelona Supercomputing Center (BSC), the Universitat Politècnica de Catalunya (UPC) and the Institut de Bioingenieria de Catalunya (IBEC) outlining the great possibilities that artificial intelligence presents to the advancement of biomaterials design and development.

The multidisciplinary team, consisting of Osnat Hakimi, Martin Krallinger and Maria Pau Ginebra, proposes to use data mining text technologies to extract information about biomaterials, which is currently dispersed across scientific articles, patents, FDA reports and congress proceedings.

These methods of advanced data mining, together with deep learning techniques, could reveal associations not previously considered between materials’ attributes and biological responses, and could help with the design and discovery of new biomaterials.

Biomaterials are materials that interact with biological systems, and are highly used in modern medicine and surgery (implants, prostheses, etc.). Their design involves tapping into complex processes, such as the interactions between cells and materials and the degradation of materials in the body.

The rising volume of published results in the field is contrasted by a low degree of sharing and systematization of data. The article explains the specific challenges in the highly multidisciplinary domain of biomaterials, and proposes steps to tackle them and enable the organization and exploitation of accumulated data.

This article has been written in the context of the DEBBIE project, a Marie Skłodowska-Curie action funded by the European Commission and dedicated to the development of the first biomaterial database using data mining tools. The project is hosted by the UPC and the BSC and can be viewed here: https://github.com/ProjectDebbie.

Article: “Time to kick- start text mining for Biomaterials” https://www.nature.com/articles/s41578-020-0215-z

Authors: Osnat Hakimi, Martin Krallinger and Maria-Pau Ginebra

DOI: https://doi.org/10.1038/s41578-020-0215-z