that was my first thought, something up with the contigs file. (full disclosure, I am involved in this project and had no problem running picard post BWA-mem when I was using the trinity.fasta file) These contigs are really ORFs that were generated from a subset of the Trinity contigs, using EMBOSS)
Duplicate @Sq Lines In Sam File Header
Hi,
I'm working on RNA-seq data and have aligned my reads to the assembled transcript using bwa-mem. When I tried to run Picard Tools (SamFormatConverter.jar), the SAM dictionary threw an exception
"Cannot add sequence that already exists in SAMSequenceDictionary".
I then checked for duplication in the SAM file header and sure enough, two entries were duplicated. How do I correct this? Commands used are given below:
bwa index contigs.fa
bwa mem contigs.fa reads_1.fq reads_2.fq >samFile.sam
java -Xmx4g -jar SamFormatConverter.jar INPUT=samFile.sam OUTPUT=bamFile.bam
This is where I faced the error.
I checked for duplicates using:
samtools view -HS samFile.sam | sort | uniq -c
2 @SQ entries were present >1 times.
Any help would be appreciated, thanks!
1 answer
The @SQ headers that bwa emits come directly from the reference sequences in the index you're mapping against.
So the likely explanation for your problem is that the same entries are duplicated in your contigs.fa file. You can check by looking at the > headers in your fasta file:
grep '>' contigs.fa | sort | uniq -c
(If there are header lines like >foo blah and >foo blah blah then these will lead to duplicate @SQ SN:foo SAM headers while not being shown as exact duplicate lines by that uniq command. So you may have to be a little careful in interpreting the results of that command.)
• 95 views
•
link
updated
by
Ram
•
written
by
John Marshall
•
written
by
John Marshall
Log in to answer this question.
More posts like this
-
What is the major problem with this pipeline of SNPs analysis?
written by kgwkk2 •
-
Duplicate SAMSequenceDictionary in SAM file - how to modify SAM header?
written by kat •
-
Exception in thread "main" htsjdk.samtools.SAMFormatException: Error parsing SAM header. @RG line m…
written by valopes •
-
Cuffdiff Error - Multi Genomic BAM not sorted.
written by Yaseen Ladak •
I do not know whether it's right nor not, but when using bwa mem, you do not use
-Mparameter, Mark short split hits as secondary (which for Picard compatibility), may be you should try this before go to picard.I think that's the problem! I'll run bwa mem again with the
-Moption. Thank you!Hi, A fresh run with the
-Moption turned on still gives the error when I try Picard Tools! :-(