A major roadblock towards accurate interpretation of single cell RNA-seq data is large technical noise resulted from small amount of input materials. The existing methods mainly aim to find differentially expressed genes rather than directly de-noise the single cell data. We present here a powerful but simple method to remove technical noise and explicitly compute the true gene expression levels based on spike-in ERCC molecules.

Availability and implementation: The software is implemented by R and the download version is available at http://wanglab.ucsd.edu/star/GRM.

Contact: wei-wang@ucsd.edu

Supplementary information: Supplementary data are available at Bioinformatics online.