Bowtie error when mapping ABI RNA-seq data with Tophat

[HSV-1]

Hi all,
I got errors like below:

No,1
=============
2012-07-26 16:37:43] Beginning TopHat run (v2.0.4)
-----------------------------------------------
[2012-07-26 16:37:43] Checking for Bowtie
Bowtie version: 0.12.8.0
[2012-07-26 16:37:44] Checking for Samtools
Samtools version: 0.1.13.0
[2012-07-26 16:37:44] Checking for Bowtie index files
[2012-07-26 16:37:44] Checking for reference FASTA file
Warning: Could not find FASTA file hg19_c.fa
[2012-07-26 16:37:44] Reconstituting reference FASTA file from Bowtie index
Executing: /home/xiaoyu/bin/bowtie-inspect hg19_c > VVinfect0h/tmp/hg19_c.fa
[2012-07-26 16:42:49] Generating SAM header for hg19_c
format: fastq
quality scale: phred33 (default)
[2012-07-26 16:43:42] Reading known junctions from GTF file
[2012-07-26 16:43:54] Preparing reads
left reads: min. length=50, max. length=50, 49053373 kept reads (712587 discarded)
[2012-07-26 17:00:17] Creating transcriptome data files..
[2012-07-26 17:01:36] Building Bowtie index from hg19UCSC.fa
[2012-07-26 17:53:34] Mapping left_kept_reads to transcriptome hg19UCSC with Bowtie
[FAILED]
Error running bowtie:
Too few quality values for read: 460T3#
are you sure this is a FASTQ-int file?
Command: /home/xiaoyu/bin/bowtie -q -C --col-keepends -v 1 -k 60 -m 60 -S -p 2 --sam-nohead --max /dev/null VVinfect0h/tmp/hg19UCSC -


No2
==============


[2012-07-26 16:43:15] Beginning TopHat run (v2.0.4)
-----------------------------------------------
[2012-07-26 16:43:15] Checking for Bowtie
Bowtie version: 0.12.8.0
[2012-07-26 16:43:15] Checking for Samtools
Samtools version: 0.1.13.0
[2012-07-26 16:43:15] Checking for Bowtie index files
[2012-07-26 16:43:15] Checking for reference FASTA file
Warning: Could not find FASTA file hg19_c.fa
[2012-07-26 16:43:15] Reconstituting reference FASTA file from Bowtie index
Executing: /home/xiaoyu/bin/bowtie-inspect hg19_c > VVinfect4h/tmp/hg19_c.fa
[2012-07-26 16:48:16] Generating SAM header for hg19_c
format: fastq
quality scale: phred33 (default)
[2012-07-26 16:48:22] Reading known junctions from GTF file
[2012-07-26 16:48:33] Preparing reads
left reads: min. length=50, max. length=50, 34854552 kept reads (1262790 discarded)
[2012-07-26 16:59:38] Creating transcriptome data files..
[2012-07-26 17:01:05] Building Bowtie index from hg19UCSC.fa
[2012-07-26 17:51:34] Mapping left_kept_reads to transcriptome hg19UCSC with Bowtie
[FAILED]
Error running bowtie:
Too few quality values for read: 51300T13
are you sure this is a FASTQ-int file?
Command: /home/xiaoyu/bin/bowtie -q -C --col-keepends -v 1 -k 60 -m 60 -S -p 2 --sam-nohead --max /dev/null VVinfect4h/tmp/hg19UCSC -


How to fix it? Is it possible to trimm the reads "460T3#" "51300T13" out? If yes, how? Please help.

[goudurix]

This happen to me one time when using -C with an illumina dataset. Are you sure your reads are in color space ?
Cheers

[HSV-1]

This happen to me one time when using -C with an illumina dataset. Are you sure your reads are in color space ?
Cheers

The data is from ABI-Solid. And I open the data file there are no ACGTs,but 1,2,3,...
They are color space.

[sonia.bao]

Got the same error here when feeding tophat2 with csfastq files as the input.

I checked the csfastq file, and found nothing wrong with it. No truncated reads or qual values. Then I tried bowtie to align the reads to the reference genome using the csfastq file as the input - bowtie finished without any error, and over 90% of the reads were mapped.

Try feeding tophat2 with csfasta+qual files as the input instead of csfastq. I tried that and tophat2 ran through successfully.

[HSV-1]

Got the same error here when feeding tophat2 with csfastq files as the input.

I checked the csfastq file, and found nothing wrong with it. No truncated reads or qual values. Then I tried bowtie to align the reads to the reference genome using the csfastq file as the input - bowtie finished without any error, and over 90% of the reads were mapped.

Try feeding tophat2 with csfasta+qual files as the input instead of csfastq. I tried that and tophat2 ran through successfully.

Thanks for your reply. How to get csfasta files and qual files from the same csfastq?

[sonia.bao]

Try this python script - it takes color space .fastq file as the input and outputs 2 files, .csfasta and .QV.qual.

(It was not written by me. Someone wrote this script and shared it on this board (much appreciated!!!). If anybody knows who the author is, please let me know and I'll update it)

csfastq2solid.py

import sys
fq = sys.argv[1]

base = fq.split(".fastq")[0]
quals = open(base + ".QV.qual", "w")
seq = open(base + ".csfasta", "w")

for i, line in enumerate(open(fq)):

    mod = i % 4
    if mod == 0: # name
        assert line[0] == "@"
        quals.write(">" + line[1:])
        seq.write(">" + line[1:])
    elif mod == 1: # cseq
        seq.write(line)
    elif mod == 3:
        print >>quals, " ".join((str(ord(q) - 33) for q in line.rstrip("\r\n")))

seq.close(); quals.close()
print >>sys.stderr, "wrote %s, %s" % (quals.name, seq.name)