This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Print Differentially Expressed Exons From Dexseq Results

Hello, I'm trying to do some differential expressed exons analysis with DEXSeq package. I got DEXSeq results by :

dxr1 = DEXSeqResults( dxdEFC, padj>=0.1 )

Then, I'm looking for print only Up-regulate or Down-regulate exons significantly expressed.

Therefore, I done :

write.table(dxr1, "DEXSeq_results" , sep="\t", col.names=T, row.names = T)

This script gave me all genes and exons in a single file within more than 60.000 rows what is very difficult to open and then difficult to sort by padj or FDR.

I think that someone can help me here to resolve that, I will be very grateful.

Best regards.

dexseq rna-seq
• 4,354 views • link
written by justinkablan225 • justinkablan225

Hello,

After printing result, I see that some same genes names are listed again in the table with another exonID. Then I've too much genes and exons, that I want is to group same geneID in one wich would group all his exons.

• 0 views • link
updated by lieven.sterck lieven.sterck • written by justinkablan225 • justinkablan225

Did you mean something like the following?

  groupID     featureID_by_groupID         
1 FBgn0000256 E009,E010,E013      
2 FBgn0000578 E009                
3 FBgn0002921 E012,E015           
4 FBgn0010909 E010                
5 FBgn0020309 E007                
6 FBgn0027579 E001,E002
• 0 views • link
written by AK AK

It can be generated by:

dxr1.sig <- as.data.frame(dxr1[dxr1$padj < 0.1 & !is.na(dxr1$padj),])
library(dplyr)
dxr1.sig.genes <- dxr1.sig %>%
  select(groupID, featureID) %>%
  group_by(groupID) %>%
  mutate(featureID_by_groupID = paste(featureID, collapse = ",")) %>%
  select(-featureID) %>%
  unique()
• 0 views • link
written by AK AK

Hello, It works but I think somethings wrong looking the results table.

groupID                 exonBaseMean       dispersion      stat             pvalue              padj                    featureID_by_groupID
ENSG00000007202 30.08387115    0.007192981     24.07148254      9.28E-07        0.000566869      E041,E043
ENSG00000007202 5.781795849    0.021773337     12.67383165      0.000370809     0.057097567      E041,E043

The same gene is listed again with different value of dispersion, pvalue and so one. This is an abstract of the results, lot of gene are the same kind of result

I don't inderstand why.

• 0 views • link
updated by lieven.sterck lieven.sterck • written by justinkablan225 • justinkablan225

Those are different exons as you're testing Differentially Expressed Exons?

• 0 views • link
written by AK AK

Then why are you simply not doing gene expression values in stead of exons? or is this part of the confusion? yes the gene expression is also taking into account the exons but then joined to an mRNA level and thus not keeping them as separate exons

• 0 views • link
written by lieven.sterck lieven.sterck

2 answers

Try:

dxr1.sig <- as.data.frame(dxr1[dxr1$padj < 0.1 & !is.na(dxr1$padj),])

then save dxr1.sig using your write.table.

• 0 views • link
written by AK AK

since you have gotten to the point where you have tab-delineated file with your values, use linux sort or even better awk. Those are applications that allow for easy and powerful manipulation (sorting, filtering on values, ... ) of tab files.

• 0 views • link
written by lieven.sterck lieven.sterck

Log in to answer this question.

More posts like this