Merging FastQ files from different runs into one?

[Zapages]

Hi guys,

I have simple small question. I was wondering if it would be fine to merge 4 runs into one fastq.gz file.

The 4 fastq files have one SRR: SRA010305

There are 4 fastq files on http://trace.ddbj.nig.ac.jp/DRASearc...?acc=SRX014987 and http://www.ebi.ac.uk/ena/data/view/SRX014987

I will use the the following code:

cat file 1.fastq 2.fastq 3.fastq 4.fastq > merged.fastq

or should I run tophat separately on each of the files.

I am looking for differential of expression between the different organs.

Thank you in advance.

[yueluo]

I think you should run tophat(and/or any other aligner) seperately for each sample. This will create an alignment file(SAM/BAM format) for each sample.
Then use other programs for DEG profiling. Since there are so many methods to do that, you might as well do some more in-forum searching

[Zapages]

Thank you for the fast reply. That sounds great.

I did some trimming of these data sets separately and did it combined. It was just to do a compare and contrast. I will now go with doing it separately. Anyway, it seems that I am able to get 16 nucleotide length for the samples. This seems awfully short to use tophat or is it fine?

Thank you again.

PS: I was using trimmomatic to trim the data sets.

[Tlexander]

Yes. Yueluo is quite right. Don't merge fastq files. Alignment them separately and then merge BAM or SAM files. It will take less time since alignment can be accelerated with separate Fastq.