Question

QuasR: Loading BAM files as start point

0

Entering edit mode

rahel14350 ▴ 10

@rahel14350-8472

Last seen 5.1 years ago

United States

Dear Dr. Stadler,

I am trying to start QuasR package with the RNA-seq BAM files I already have from paired end reads. This is my sample_file.txt:

FileName   SampleName
C_mapped.bam   Sample1
D_mapped.bam   Sample2

> proj2 <- qAlign(sampleFile, genomeFile, splicedAlignment=TRUE, clObj=cl)
Error: samples_rna_paired_bam.txt should contain the column names FileName1,FileName2,SampleName (paired end data)

and when I add paired status:

> proj2 <- qAlign(sampleFile, genomeFile, splicedAlignment=TRUE, paired='fr', clObj=cl)
Error in createQProject(sampleFile, genome, auxiliaryFile, aligner, maxHits, :
object 'fr' not found

Would you please help me with this error?

Many thanks in advance,

Rahel

quasr • 2.1k views

ADD COMMENT • link updated 8.6 years ago by Michael Stadler ▴ 350 • written 8.6 years ago by rahel14350 ▴ 10

score 1 · Answer 1 · 2015-09-02

1

Entering edit mode

Michael Stadler ▴ 350

@michael-stadler-5887

Last seen 2.3 years ago

Switzerland

Dear Rahel,

My guess is that you have a problem in the file "sample_file.txt". Make sure that it only contains two columns in each row, separated by a single tab-character. The first error message that you got ("... should only contain the columns names ...") is only thrown when your samples file contains three fields (two tab-characters) in a row.

Also, as you have already guessed, you will need to tell qAlign the paired status of your bam files, using paired="fr" as you have tried in your second call. I am not sure why you get the "object 'fr' not found" error message there. Maybe you forgot to put fr into (double-)quotes?

Regards,

Michael

ADD COMMENT • link 8.6 years ago Michael Stadler ▴ 350

0

Entering edit mode

Dear Michael,

Many thanks for your prompt reply. You were right, I had two tab in my sample file txt file, when I edit it to one-tab file, I had no error with qAlign.

I have three more question, I have a big Stranded RNA-seq data set with many BAM files that are produced by TopHat2 and I am looking for intron retention and exon skipping in the BAM files. Is it still reasonable to use splicedAlignment=TRUE in the qalign command, As I did not use splicediff for mapping?

And Can I say the KnownIntrons which coming after exonJunctions count are referring to the intron retention in my data? What about exon skipping, Are they included in the exonBodyLevels?

Many thanks in advance,

Kind Regards,

Rahel

ADD REPLY • link 8.6 years ago rahel14350 ▴ 10

0

Entering edit mode

Dear Michael,

I have a new question. Is that possible to link the "exonjunctions" output to the related gene IDs. I could do this for the exon levels. But is there any way for the Junctions as well?

Kind Regards,

Rahel

ADD REPLY • link 8.6 years ago rahel14350 ▴ 10

score 1 · Answer 2 · 2015-09-03

Dear Rahel,

Regarding splicedAlignment=TRUE: As you are using external bam files, this parameter will not have any effect on the alignments themselves. It would only matter if you are doing the alignments within QuasR. It's purpose is here to "tell" QuasR that the bam files may contain spliced alignments.

Regarding qCount(..., reportLevel="junctions"), this will detect and count alignment support for all intron-like gaps in your alignments (deletions in the read relative to the reference of at least 60bp). It is not meant to classify splicing events like retained introns or skipped exons. Such events could be detected on the basis of junction and exon body counts obtained using qCount, but this detection is not covered within QuasR.

All of this is covered in the vignette.

Michael