Changes between Version 23 and Version 24 of SOPs/rna-seq-diff-expressions


Ignore:
Timestamp:
08/06/15 09:59:16 (10 years ago)
Author:
thiruvil
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • SOPs/rna-seq-diff-expressions

    v23 v24  
    7474       * Be careful with interpretation of images, as they appear to show that upper-quartile normalization does more bad than good.
    7575       * Sample commands:
    76          * ''# Normalize to the 75th percentile''
    77          * ''R --vanilla < normalize_DGE_matrix.R DGE_sample_input.txt percentile=0.75 DGE_sample_input.percentile.0.75.txt''
    78          * ''# Normalize to the mean''
    79          * ''R --vanilla < normalize_DGE_matrix.R DGE_sample_input.txt mean DGE_sample_input.mean.txt''
     76         * ''Normalize to the 75th percentile''
     77           * ''R --vanilla < normalize_DGE_matrix.R DGE_sample_input.txt percentile=0.75 DGE_sample_input.percentile.0.75.txt''
     78         * ''Normalize to the mean''
     79           * ''R --vanilla < normalize_DGE_matrix.R DGE_sample_input.txt mean DGE_sample_input.mean.txt''
    8080       * Calculating library size
    8181         * We can do this the most obvious way using the total number of reads in each sample (column of our counts matrix), which is shown below.  Many methods for differential expression analysis use more complex methods to calculate what they refer to as something like "effective library size" (such as DESeq's estimateSizeFactors() command), which appears to work better than the more obvious method.