genome Space to transcriptome space

[ashokrags]

hello,
I wonder if anybody else has come across this issue or has some insight into the same. I have RNA-seq data and I usually align to the genome using GSNAP or Tophat. Thus far I have found Cufflinks to be the only major tool to estimate Isoform FPKMs using this approach (I know there are a couple of other tools out there). For comparison, I wanted to estimates FPKMs using RSEM, eXPress or BitSeq with the same bams, but am stuck as they require the alignments to be made to the reference transcriptome not the genome?? Any suggestions to circumvent this issue other than complete realignment??. Some thoughts along this line are :

1. Is there a tool to convert genome space to transcriptome space and vice versa

1. would such a tool be useful

1. what would be needed in this context?

1. sam flags when reads map to the same exons in two isoforms?

1. will realignment be needed?

Any thoughts, insights in this context will be much appreciated
cheers
Ashok

[GenoMax]

Something like this has been done for the TCGA RNA-seq data (see #10): https://webshare.bioinf.unc.edu/publ...eq_summary.pdf

The software (UBU) is available here (sam-xlate - Translate from genome to transcriptome coordinates) : https://github.com/mozack/ubu/wiki

[rskr]

hello,
I wonder if anybody else has come across this issue or has some insight into the same. I have RNA-seq data and I usually align to the genome using GSNAP or Tophat. Thus far I have found Cufflinks to be the only major tool to estimate Isoform FPKMs using this approach (I know there are a couple of other tools out there). For comparison, I wanted to estimates FPKMs using RSEM, eXPress or BitSeq with the same bams, but am stuck as they require the alignments to be made to the reference transcriptome not the genome?? Any suggestions to circumvent this issue other than complete realignment??. Some thoughts along this line are :

1. Is there a tool to convert genome space to transcriptome space and vice versa

1. would such a tool be useful

1. what would be needed in this context?

1. sam flags when reads map to the same exons in two isoforms?

1. will realignment be needed?

Any thoughts, insights in this context will be much appreciated
cheers
Ashok

I thought cufflinks would do something similar. Maybe generate a transcriptome fasta file that could then be re-mapped. I think a tool like this would be very useful though, since working in the genome space is overkill when the mapped regions are much smaller, also in many cases it makes sense to think about the transcripts as contiguous.