Should Artificial Intelligence Play a Durable Role in Biomedical Research and Practice?

Abstract

During the last decade, artificial intelligence (AI) was applied to nearly all domains of human activity, including scientific research. It is thus warranted to ask whether AI thinking should be durably involved in biomedical research. This problem was addressed by examining three complementary questions (i) What are the major barriers currently met by biomedical investigators? It is suggested that during the last 2 decades there was a shift towards a growing need to elucidate complex systems, and that this was not sufficiently fulfilled by previously successful methods such as theoretical modeling or computer simulation (ii) What is the potential of AI to meet the aforementioned need? it is suggested that recent AI methods are well-suited to perform classification and prediction tasks on multivariate systems, and possibly help in data interpretation, provided their efficiency is properly validated. (iii) Recent representative results obtained with machine learning suggest that AI efficiency may be comparable to that displayed by human operators. It is concluded that AI should durably play an important role in biomedical practice. Also, as already suggested in other scientific domains such as physics, combining AI with conventional methods might generate further progress and new applications, involving heuristic and data interpretation.

Keywords: deep learning; immunology; machine learning; omic; statistics.

Figure 1

Early mechanisms of inflammation. A blood leukocyte flowing with a velocity of several hundreds of µm/s in a capillary blood vessel (0) is bound by selectin molecules expressed on the membrane of endothelial cells subjected to an inflammatory stimulus (1). This results in a nearly hundredfold velocity reduction, leading to a jerky motion called rolling (2). Cytokines bound to the membrane of activated endothelial cells then activate the surface integrins of rolling leukocytes, leading to a firm attachment to cell adhesion molecules expressed by endothelial cells (3). The leukocyte will then transmigrate towards surrounding tissues (4, 5, 6).

Figure 2

TCR-mediated T-lymphocyte activation. The figure shows a minimal view of molecules participating in the initial step of T-lymphocyte activation: An antigen presenting cell (APC) displays tens of thousands of major histocompatibility molecules (MHC) bearing various oligopeptides (p). The T cell receptor (TCR) is associated with the so-called CD3 complex typically made of three dimers (γε, δε, ζζ), the cytoplasmic part of which contains tyrosine-bearing motives (immunoreceptor tyrosine-based activation motifs called ITAMs, shown as colored ellipses). The association between a TCR and its cognate pMHC ligand enhances the binding of a co-receptor (such as CD8 on the figure) that is often associated with a tyrosine kinase (lck). TCR-pMHC interaction will thus trigger the phosphorylation of ITAMs by lck, and phosphorylated ITAMs act as docking sites for another kinase (zeta-associated protein of 70 kDa, ZAP70). Note this is a very simplified view. More details may be found elsewhere [38].