TopHat - Cufflinks problems in a multifasta file

ccstaats · 01-31-2012, 04:35 AM

Hello all
I'm new to rnaseq analysis, so I wonder if you guys can help me with some issues.
So, I'm trying to evaluate differential expression from the WT and a mutant strain of a yeast, whose genome was sequenced at Broad Institute.
In a first attempt, I'm trying only to evaluate the genes that have been annotated.
The format that comes from Broad is a fasta file with up to 10.000 genes. The genome is not annotated.
In this way, I've downloaded the fasta file with the genes and started to wor on it.
First, I've build a database using bowtie-build.
Then I mapped the reads to the db using tophat
(tophat -r 30 -o /Users/Analises/db/tophat/delta genesmod4.fas 101.fastq) It runs fine
Bou when I tried to run cufflins (cufflinks -o /Users/Analises/db/cufflinks_delta/ /Users/Analises/db/tophat/delta/accepted_hits.bam), I've got this message:

You are using Cufflinks v1.3.0, which is the most recent release.
[10:06:58] Inspecting reads and determining fragment length distribution.
> Processing Locus ABC0001 [ ] 0%
Error: this SAM file doesn't appear to be correctly sorted!
current hit is at ABC_0002:3, last one was at ABC_0001:5679
Cufflinks requires that if your file has SQ records in
the SAM header that they appear in the same order as the chromosomes names
in the alignments.
If there are no SQ records in the header, or if the header is missing,
the alignments must be sorted lexicographically by chromsome
name and by position.

Any clues?
Thank you in advance

Charley

jbrwn · 01-31-2012, 10:08 AM

there may be problems in other areas, but it's difficult to troubleshoot with the given information. things that would help are fragment and read length, whether the reads are actually paired-end or not.

for the final problem though, it's likely as simple as it suggests. just sort the output using:
samtools sort aln.bam aln.sorted

see: http://samtools.sourceforge.net/samtools.shtml

ccstaats · 01-31-2012, 10:51 AM

Hello jbrwn
Thank you for your answer
I think I've found a way to escape away from this problem
I was only considering the transcript fasta file from the Broad Institute.
Know, I got the whole genome fasta file and the gft.
Alll the pathway from tophat to cuffdiff was simple.
Thank you very much
Charley

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
cufflinks problems	wall_y	Bioinformatics	3	08-06-2013 09:31 PM
Best source for GTF file for use with TopHat/Cufflinks	sdarko	Bioinformatics	17	12-14-2012 12:48 PM
Problem mapping to multifasta reference file	Mamoon Rashid	Bioinformatics	0	05-14-2011 09:19 AM
Cufflinks -G Problems	ayfchang	Bioinformatics	1	02-10-2011 05:59 PM
Tophat/Cufflinks using own GFF3 file problems	James	Bioinformatics	5	07-04-2010 11:55 AM

01-31-2012, 04:35 AM	#1
ccstaats Member Location: Porto Alegre Join Date: Jan 2012 Posts: 12	TopHat - Cufflinks problems in a transcript multifasta file whitout coordinates Hello all I'm new to rnaseq analysis, so I wonder if you guys can help me with some issues. So, I'm trying to evaluate differential expression from the WT and a mutant strain of a yeast, whose genome was sequenced at Broad Institute. In a first attempt, I'm trying only to evaluate the genes that have been annotated. The format that comes from Broad is a fasta file with up to 10.000 genes. The genome is not annotated. In this way, I've downloaded the fasta file with the genes and started to wor on it. First, I've build a database using bowtie-build. Then I mapped the reads to the db using tophat (tophat -r 30 -o /Users/Analises/db/tophat/delta genesmod4.fas 101.fastq) It runs fine Bou when I tried to run cufflins (cufflinks -o /Users/Analises/db/cufflinks_delta/ /Users/Analises/db/tophat/delta/accepted_hits.bam), I've got this message: You are using Cufflinks v1.3.0, which is the most recent release. [10:06:58] Inspecting reads and determining fragment length distribution. > Processing Locus ABC0001 [ ] 0% Error: this SAM file doesn't appear to be correctly sorted! current hit is at ABC_0002:3, last one was at ABC_0001:5679 Cufflinks requires that if your file has SQ records in the SAM header that they appear in the same order as the chromosomes names in the alignments. If there are no SQ records in the header, or if the header is missing, the alignments must be sorted lexicographically by chromsome name and by position. Any clues? Thank you in advance Charley Last edited by ccstaats; 01-31-2012 at 04:50 AM.

01-31-2012, 10:08 AM	#2
jbrwn Member Location: Denver, CO Join Date: Mar 2011 Posts: 37	there may be problems in other areas, but it's difficult to troubleshoot with the given information. things that would help are fragment and read length, whether the reads are actually paired-end or not. for the final problem though, it's likely as simple as it suggests. just sort the output using: samtools sort aln.bam aln.sorted see: http://samtools.sourceforge.net/samtools.shtml

01-31-2012, 10:51 AM	#3
ccstaats Member Location: Porto Alegre Join Date: Jan 2012 Posts: 12	Hello jbrwn Thank you for your answer I think I've found a way to escape away from this problem I was only considering the transcript fasta file from the Broad Institute. Know, I got the whole genome fasta file and the gft. Alll the pathway from tophat to cuffdiff was simple. Thank you very much Charley