Gene expression is first normalized to account for total expression within each cell, then a log transform is applied. To select genes for downstream unsupervised analyses (e.g. PCA, t-SNE), the expression values are corrected for heteroscedasticity using the variance stabilizing transformation (VST). We also apply a method to regress out the per-cell measures of total UMI counts, total genes detected, and percentage of reads mapping to mitochondria.
How do you account for technical variability? Does the program fit for mean variance trend, test for non zero biological variability, or deconvolution based normalization? Print
Modified on: Mon, 26 Oct, 2020 at 11:01 AM
Did you find it helpful? Yes No
Send feedbackSorry we couldn't be helpful. Help us improve this article with your feedback.