Tophat~Error retrieving prep_reads info

[ruc9]

Hi, everyone.

I got the problem below when I was running tophat on a RNAseq data.
The error message are below:

rui@:~/RNAseq$ tophat -p 20 --solexa-quals ~/reference/bowtie_ref/hg19bowtieidx /home/NGS/Sample_RNAseq49T/RNAseq49T_NoIndex_L001_R1_001.fastq /home/NGS/Sample_RNAseq49T/RNAseq49T_NoIndex_L001_R2_002.fastq

[Tue Jan 24 13:34:59 2012] Beginning TopHat run (v1.4.0)
-----------------------------------------------
[Tue Jan 24 13:34:59 2012] Preparing output location ./tophat_out/
[Tue Jan 24 13:34:59 2012] Checking for Bowtie index files
[Tue Jan 24 13:34:59 2012] Checking for reference FASTA file
[Tue Jan 24 13:34:59 2012] Checking for Bowtie
Bowtie version: 0.12.7.0
[Tue Jan 24 13:34:59 2012] Checking for Samtools
Samtools Version: 0.1.18
[Tue Jan 24 13:34:59 2012] Generating SAM header for /home/rui/reference/bowtie_ref/hg19bowtieidx
format: fastq
quality scale: solexa33 (reads generated with GA pipeline version < 1.3)
[Tue Jan 24 13:35:02 2012] Preparing reads
left reads: min. length=100, count=3964145
[FAILED]
Error retrieving prep_reads info.

Does anybody know what happened?

Thanks a lot.

Rui

[Ambily]

Did you figure a way around this error ?

I have also been getting the same error for some few cases while the rest of them ran without any issues.

I would appreciate any pointers regarding this.

[ruc9]

Hi, Ambily.

I sent email to the person who write Tophat. And I went back to check my fastq files.

Since the output still print out the left reads information, my problem is the second reads file is corrupted. It is almost all N's. And this problem may also due to the file format is incorrect. Hope this will give you some help.

Rui

[Ambily]

Thank you so much, Rui. I will look into my files once again and see what's wrong.

Ambily

[Ambily]

Hi Rui,

I looked at my reads and, as you had mentioned, mine contains too many Ns too. So, could you please advise me as to what needs to be done ? Is there some workaround, or should we repeat the experiment itself ?

Thanks in advance.

[ruc9]

Hi, Ambily.

I don't think we can workaround since reads contain too many Ns, even if we can do something, the result may not be reliable. Our lab decide to redo the experiment. I suggest you try to ask somebody else like the person who wrote Tophat, since I am also new to RNAseq analysis.

Rui

[Ambily]

Thank you so much, Rui. I will check up with Tophat authors too before I make a final decision. Highly appreciate your prompt response.