Evaluation of a computational model addressing synaptic pruning defects and their effect on working memory in autism spectrum disorder

Abstract

Synaptic pruning contributes to the formation of “small-world nature” in brain networks. In autism spectrum disorder (ASD), abnormalities in the related gene AUTS2 are thought to disappear this property, resulting in working memory (WM) dysfunction. This study, we constructed ASD and typical development (TD) models using the Watts-Strogatz model and evaluated WM function by means of a 2-back task. The results revealed that the correct answer rate of the ASD model was lower than that of the TD model. This suggests that the loss of “small-world nature” in the brain network reduces the efficiency of information propagation, leading to degradation of WM ability.