Allows users to capture and visualize the low-dimensional structures in single-cell gene expression data. scvis is a robust latent variable model that allows to spot underlying low-dimensional structures in scRNA-seq data. It learns a parametric mapping from the high-dimensional space to a low-dimensional embedding. This tool estimates the uncertainty of mapping a high-dimensional point to a low-dimensional space which adds rich capacity to interpret results.
Department of Computer Science, University of British Columbia, Vancouver, BC, Canada; Department of Molecular Oncology, BC Cancer Agency, Vancouver, BC, Canada; Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, BC, Canada
scvis funding source(s)
Supported by a Discovery Frontiers project grant, “The Cancer Genome Collaboratory”, jointly sponsored by the Natural Sciences and Engineering Research Council (NSERC), Genome Canada (GC), the Canadian Institutes of Health Research (CIHR) and the Canada Foundation for Innovation (CFI); the BC Cancer Foundation; the Canadian Breast Cancer Foundation, the Canadian Cancer Society Research Institute (impact grant 701584) the Terry Fox Research Institute (PPG program on forme fruste tumors), CIHR (grant MOP-115170), CIHR Foundation (grant FDN-143246); the Canada Research Chairs.