Hi,
I have 3 reference annotation GTF files and 3 query GTF files which were generated by Tophat-Cufflinks combination.
I am using cuffcompare to compare the 3 query GTF files against each reference annotation GTF file.
While doing this I noticed that the resulting combined.gtf and .tracking file are not reporting all the transcripts from the query GTF files.
When I checked the .stats file I noticed in 'Summary for dataset' section that only a subset of the transcripts are reported.
Example:
GTF_A - 265244 transcripts
GTF_B - 267648 transcripts
.stats file for GTF_A reports always the same number against all references
#= Summary for dataset: GTF_A.gtf :
# Query mRNAs : 216023 in 57092 loci (188694 multi-exon transcripts)
.stats file for GTF_B also reports always the same number against all references
#= Summary for dataset: GTF_B.gtf :
# Query mRNAs : 218034 in 58267 loci (189378 multi-exon transcripts)
And even though the reference annotation is different the list of missing transcripts from query GTF is exactly the same.
So they must have been skipped when importing the transcripts from query GTF and it is independent of the reference GTF.
When comparing GTF_B (Query) vs GTF_A (Reference).
.stats file reports
#= Summary for dataset: GTF_B.gtf :
# Query mRNAs : 218034 in 58267 loci (189378 multi-exon transcripts)
# (21371 multi-transcript loci, ~3.7 transcripts per locus)
# Reference mRNAs : 238139 in 45259 loci (223605 multi-exon)
One thing to notice above is that the statistics for GTF_A (Reference) are different when it is used a reference from earlier when it was used in query.
I tried to find any trend for the missing transcripts but cannot identify anything.
Can anyone help me understand the issue at hand.
Regards,
Veer
I have 3 reference annotation GTF files and 3 query GTF files which were generated by Tophat-Cufflinks combination.
I am using cuffcompare to compare the 3 query GTF files against each reference annotation GTF file.
While doing this I noticed that the resulting combined.gtf and .tracking file are not reporting all the transcripts from the query GTF files.
When I checked the .stats file I noticed in 'Summary for dataset' section that only a subset of the transcripts are reported.
Example:
GTF_A - 265244 transcripts
GTF_B - 267648 transcripts
.stats file for GTF_A reports always the same number against all references
#= Summary for dataset: GTF_A.gtf :
# Query mRNAs : 216023 in 57092 loci (188694 multi-exon transcripts)
.stats file for GTF_B also reports always the same number against all references
#= Summary for dataset: GTF_B.gtf :
# Query mRNAs : 218034 in 58267 loci (189378 multi-exon transcripts)
And even though the reference annotation is different the list of missing transcripts from query GTF is exactly the same.
So they must have been skipped when importing the transcripts from query GTF and it is independent of the reference GTF.
When comparing GTF_B (Query) vs GTF_A (Reference).
.stats file reports
#= Summary for dataset: GTF_B.gtf :
# Query mRNAs : 218034 in 58267 loci (189378 multi-exon transcripts)
# (21371 multi-transcript loci, ~3.7 transcripts per locus)
# Reference mRNAs : 238139 in 45259 loci (223605 multi-exon)
One thing to notice above is that the statistics for GTF_A (Reference) are different when it is used a reference from earlier when it was used in query.
I tried to find any trend for the missing transcripts but cannot identify anything.
Can anyone help me understand the issue at hand.
Regards,
Veer
Comment