Takes as input "raw", integer count expression matrix of quantified sequenced reads and outputs a scale-normalised, log-transformed expression matrix. Developed for the output of htseq-count.
Version | 0.1 |
---|---|
Bundle | sequencing |
Categories | Normalization |
Authors | Katherine Icay (katherine.icay@helsinki.fi) |
Issue tracker | View/Report issues |
Requires | libxml2-dev (DEB) ; EdgeR (R-bioconductor) ; DESeq (R-bioconductor) |
Source files | component.xml normalised_exprMatrices.r |
Usage | Example with default values |
Name | Type | Mandatory | Description |
---|---|---|---|
expr | CSV | Mandatory | Expression matrix of raw counts. First column is miRNA/smallRNA IDs. Column headers are sample names. Recommended: output file of HTSeqExprMatrix . |
Name | Type | Description |
---|---|---|
array | Array<CSV> | Array of scale-normalised, log-transformed expression matrices. Two files for each option parameter are produced: one where log(0) values are replaced with NA to prevent -Inf values, and one where a pseudocount of 1 is added to the entire expression matrix to preserve the profile and ensure log(0+1) is 0. |
Name | Type | Default | Description |
---|---|---|---|
options | string | "edger" | Comma-seperated list of possible scale-normalization methods to use. Possible options are "deseq", "edgeR" and/or "uqnorm". Options are not case-sensitive, but are spelling-sensitive. DESeq = counts scaled by library size (estimateSizeFactors). EdgeR = counts per million. UQnorm = counts scaled by the library's top 25th percentile. |
Test case | Parameters▼ | IN expr |
OUT array |
|||
---|---|---|---|---|---|---|
case1 | properties | expr | array | |||
options=edger,deseq, uqNorm |