Home

Welcome to the SCAPE wiki

SCAPE (Single Cell Alternative Polyadenylation using Expectation-maximization) is a tool to explore alternative polyadenylation at single cell level based 3'-tag sequencing data.

SCAPE models the scRNA-seq data as a mixture of K isoform components and one noise component, i.e. a read either comes from an isoform or belongs to noise. The $i$th ($i$=1…K) isoform component (Fig. 1) has three parameters: (a) the mean position of the pA site $$α_i$$, (b) the standard deviation (std) of the pA site $$β_i$$, (c) the weight or proportion $$π_i$$. The number of components is automatically selected using Bayesian Information Criterion (BIC). We use the Expectation-Maximization (EM) algorithm and numerical optimization to infer the parameters.

For questions regarding the usage of SCAPE use the issues section.