Abstract
Biological motion perception is an essential ability of human visual system, which plays a major adaptive role in identifying, interpreting and predicting actions of others. Here we employed a genome-wide association method to examine the genetics of biological motion perception as a polygenetic cognitive trait. We measured the biological motion detection ability in a healthy cohort of Chinese population with normal or corrected-to-normal vision. Visual stimuli were point-light human walker figures. In the behavioral task, subjects were presented with two successive random motion dot animations and made a two-alternative forced choice (2AFC) judgment on which animation contained a point-light walker. We measured the average point-light walker detection accuracy for each subject. The behavioral performance showed substantial individual differences. We performed a genome-wide association study (GWAS) in a discovery cohort of 845 participants to identify biological-motion-detection-related variants. 125 common single nucleotide polymorphisms (SNPs) showing suggestive genomic significance (p < 10-4) were picked out for further replication in another cohort of 2102 Chinese people. 6 SNPs passed the replication study (nominal p < 0.05). Then we used functional magnetic resonance imaging (fMRI) to functionally validate these candidate SNPs in another cohort of 64 Chinese people. During viewing point-light walkers (relative to randomly moving point-light dots), the neural activity of the posterior superior temporal sulcus (pSTS) in the right hemisphere – a critical cortical area for biological motion processing – was taken as a phenotype. Results showed DGKD (nearest gene for SNP rs1053895) was associative with biological motion perception (nominal p < 0.05). Besides, SNP-based heritability was estimated to be 14.1% for biological motion detection ability. Evidence from this study indicates a mild contribution of genetics to human biological motion perception and suggests specific genes associated with it.
Meeting abstract presented at VSS 2018