Honda, N. | Search bioRxiv

bioRxiv2023

Leveraging Machine Learning to Uncover the Hidden Links between Trusting Behavior and Biological Markers

Understanding the decision-making mechanisms underlying trust is essential, particularly for patients with mental disorders who experience difficulties in developing trust. We aimed to explore biomarkers associated with trust-based decision-making by quantitative analysis. However, quantification of decision-making properties is difficult because it cannot be directly observed. Here, we developed a machine learning method based on Bayesian hierarchical model to quantitatively decode the decision-making properties from behavioral data of a trust game. By applying the method to data of patients with MDD and healthy controls, we estimated model parameters regulating trusting decision-making. The estimated model was able to predict behaviors of each participant. Although there is no difference of the estimated parameters between MDD and healthy controls, several biomarkers were associated with the decision-making properties in trusting behavior. Our findings provide valuable insights into the trusting decision-making, offering a basis for developing targeted interventions to improve their social functioning and overall well-being.

bioinformatics↗

bioRxiv2021

Adaptive discrimination of antigen risk by predictive coding in immune system

The immune system discriminates between harmful and harmless antigens based on past experiences; however, the underlying mechanism is largely unknown. From the viewpoint of machine learning, the learning system predicts the observation and updates the prediction based on prediction error, a process known as predictive coding. Here, we modeled the population dynamics of T cells by adopting the concept of predictive coding; helper and regulatory T cells predict the antigen amount and excessive immune response, respectively. Their prediction error signals, possibly via cytokines, induce their differentiation to memory T cells. Through numerical simulations, we found that the immune system identifies antigen risks depending on the concentration and input rapidness of the antigen. Further, our model reproduced history-dependent discrimination, as in allergy onset and subsequent therapy. Together, this study provided a novel framework to improve our understanding of how the immune system adaptively learns the risks of diverse antigens.

immunology↗