QWENDY: Gene Regulatory Network Inference by Quadruple Covariance Matrices
by Yue Wang, Xueying Tian
Read the paper
In this paper, we study how to infer the regulatory relations between genes, where the input is the single-cell expression levels of genes at four time points. Due to technical limitations, each cell can be measured only once. Thus we have four probability distributions without their joint distribution, and we want to solve this dynamics. Our approach is to study how the covariance matrix evolves, and solve it analytically without any optimization. Our method, named QWENDY, ranks the first among 17 methods on three experimental data sets.
QWENDY method: solving gene regulation from covariance
