Confusion regarding Trinity analyze_diff_expr.pl

Question

[pbigbig]

Hi everyone,

I have some confusion regarding two ways of usage of analyze_diff_expr.pl script in the Trinity package.

In the usage instruction resides within script, we have to input matrix file: ${name}.TMM.EXPR.matrix , as following:

 # Required:
 #
 #  --matrix <string>       TMM.EXPR.matrix
 #
 # Optional:
 #
 #  -P <float>             p-value cutoff for FDR  (default: 0.001)
 #
 #  -C <float>             min abs(log2(a/b)) fold change (default: 2  (meaning 2^(2) or 4-fold).
 #
 #  --output <float>       prefix for output file (default: "diffExpr.P${Pvalue}_C${C})
 #
  

But as I learned from other guides, some used script run_TMM_normalization_write_FPKM_matrix.pl to perform TMM normalization, as in:

/run_TMM_normalization_write_FPKM_matrix.pl -­‐-­‐matrix${name}.counts.matrix -­‐-­‐lengths trans_lengths.txt

And then use the output ${name}.counts.matrix.TMM_normalized.FPKM as the input matrix for analyze_diff_expr.pl.

I tried both matrices as input and they give quite similar heatmap pattern, despite values varied between them. Could someone please explain the different between ${name}.TMM.EXPR.matrix and ${name}.counts.matrix.TMM_normalized.FPKM ?

Thank you very much!

Answer 1

[Farbod]

Dear pbigbig, Hi.

I think it is best to ask your question from the Trinity support team which is very active.

And some result file names are depends to the Trinity version you are using;

When you ran the 'abundance_estimates_to_matrix.pl ' step, it should have created a TMM normalized fpkm matrix file along with your counts.matrix file. You can use the TMM normalized matrix from that step for your heatmap.

~ Best

[pbigbig]

thank you for your suggestion !