Error while doing EdgeR fro miRNA Seq analysis

Question

[nivya.james2016]

Hi all,

I am presently doing miRNASeq analysis. I have come across an error while doing the differential expression analysis. Could anyone please help me to rectify this error?? I am pasting the commands below;

Got 3 samples, and got: 3 data fields.
Header: miRNA ID    Early Stage Stage III
Next: hsa-miR-6829-5p   7   56

-shifting sample indices over.
$VAR1 = {
          'Early Stage' => 1,
          'Stage III' => 2
        };
$VAR1 = {
          'SRR7189569' => [
                            'Early'
                          ],
          'SRR7189579' => [
                            'Stage'
                          ]
        };
Contrasts to perform are: $VAR1 = [
          [
            'SRR7189569',
            'SRR7189579'
          ]
        ];
Error, cannot determine column index for replicate name [Early]$VAR1 = {
          'Early Stage' => 1,
          'Stage III' => 2
        };

Thank you in advance

[andrew.j.skelton73]

This doesn't look like traditional edgeR within the R environment. Can you clarify how you're executing edgeR?

[nivya.james2016]

Yes sure

This was my command for executing edgeR

/home/nivya/Downloads/Trinityrnaseq-v2.6.6/Analysis/DifferentialExpression/run_DE_analysis.pl \
--matrix /home/nivya/Desktop/NSCLC/miRNASeq/DEG/matrix.txt \
--method edgeR \
--samples_file sample \
--dispersion 0.1 \
--output edgeR_dir

[Devon Ryan]

Why are you using Trinity? If you want to use edgeR, just use edgeR.

[h.mon]

Trinity has a plethora of helper scripts, wrapping all kinds of analyses. Although they can be helpful if they work properly, if they fail, they only add complexity to the debugging task.

@nivya.james2016: the script is trying to assign which counts matrix columns should be assigned to each treatment. Apparently you used spaces in counts matrix column names and / or treatments, my guess is these spaces are causing incorrect parsing of names. Try replacing the spaces by underscores, or just remove them.