fixed directionality assignment

dge_workflow/featcounts_deseq.R

@@ -257,7 +257,7 @@ if(!is.null(qcutoff)){
 }
 #deResults %<>% mutate(is_hit=pvalue<0.05)
 
-deResults %<>% mutate(s1_overex=s1_over_s2_logfc>1)
+deResults %<>% mutate(s1_overex=s1_over_s2_logfc>0)
 
 
 normCounts <- counts(dds,normalized=TRUE) %>%

[brandl]

@lakshman I hope that this patch will fix the problem. I've reprocessed Lenka's data and around 10% of the data were affected (genes with a fold-change of less than 1 which were still called as being differentially expressed)

Please let me know if you spot other major differences compared to cuffdiff

[lakshman]

@brandl I tried to rerun line 99 from sarita_flyome.sh but i get the following pandoc error

[brandl]

pandoc is not in your PATH but is required by rmarkdown. On madmax (and indirectly via the mount-point also on bioinfo) this one would work /lustre/projects/bioinfo/holger/bin/pandoc-1.13.1/linux/rpm/x86_64/pandoc

[lakshman]

@brandl I tried this export PATH=/projects/bioinfo/holger/bin/pandoc-1.13.1/linux/rpm/x86_64/pandoc:$PATH but the same error persists

[brandl]

you'd need to export export PATH=/projects/bioinfo/holger/bin/pandoc-1.13.1/linux/rpm/x86_64:$PATH instead I think.

[lakshman]

@brandl Thanks. it worked. Now I get this error about the plot:

I had a look at the yintercept parameter but it seems fine to me. any ideas? Thanks