Dr. Hongzhe Li and his AMCS PhD student, Haoshu Xu, have published a new machine learning method in the Journal of Machine Learning Research (JMLR) to facilitate population-scale single-cell data analysis.
Their method addresses the challenging problem of regression analysis for covariance matrix–valued outcomes with Euclidean covariates, motivated by the need to model personalized gene co-expression matrices. Applying this framework to large-scale human PBMC single-cell data, the researchers uncovered age-related dysregulation in gene co-expression networks, particularly among genes involved in nutrient-sensing pathways, especially a loss of co-expression among these genes after the age of 60.