-
Notifications
You must be signed in to change notification settings - Fork 9
User Guide
mdelmans edited this page Jan 3, 2019
·
6 revisions
To run D3E analysis please run D3ECmd.py:
python D3ECmd.py InputFile OutputFile Label1 Label2 [-h] [-m {0,1}] [-t {0,1,2}] [-z {0,1}] [-n {1,0}] [-s {1,0}] [-v]
Mandatory arguments:
- InputFile : path to the read-count table
- OutputFile : path to the output file
- Label1 : a common label for the cells of the first type
- Label2 : a common label for the cells of the second type
Optional arguments
- -m (--mode) : run mode (default = 1)
- -test (--test) : test for distribution comparison
- -z (--removeZeros) : if -z is set to 1, all zero entities in the read-count table will be removed prior to analysis (default = 0)
- -n (--normalise) : if -n is set to 1, a normalization routine will be performed before analysis (default = 1)
- -v (--verbose) : run in verbose mode (default)
D3ECmd.py accepts a tab-separated read-count table, where rows correspond to genes, and columns correspond to individual cells. The file should have a header row which has the following tab-separated format:
"GeneID Label<sub>1</sub> Label<sub>2</sub> Label<sub>3</sub> ... "
where Li are the cell type labels. Differential expression analysis can be performed on two cell types at a time.
Each line should start with a gene ID, followed by a sequence of read-counts. Empty lines are ignored.
D<sup3 can run in two modes, specified by -m (--mode) option:
- Mode 0 : Methods of moments is used to estimate parameters ( fast but less accurate )
- Mode 1 : Bayesian inference is used to estimate parameters ( slow but more accurate, default )
D<sup3 uses one of the following distribution comparison tests, specified by -t (--test) option:
- Mode 0: Cramer-von Mises test
- Mode 1: Kolmogorov-Smirnov test
- Mode 2: Anderson-Darling test
- Mode 3: Likelihood ratio test
D3E produces a table as an output with the following columns:
- Gene id : ID of a gene, that matches ID in the input file.
- a, b, g : Parameter values of the fitted transcriptional bursting model.
- gof : Goodness of transcriptional bursting model fit.
- s, f, d : Average burst size, expression frequency and duty cycle of a gene.
- Rs, Rf, Rd : Log2 fold-change of the corresponding parameters.
- p-value : p-value for the null-hypothesis, that two genes are not differentially expressed.
- μ, cv : mean and coefficient of variation of the expression level in each corresponding cell type.