Concatenate several SRA reads to a single fastq file

[krespim]

Hi all,

I want to re-analyse a dataset available in GEO:

I've downloaded all the files, however it seems that each replicate/experiment has several file corresponding to several runs. The question is: how to concatenate these runs once they are converted into fastq? Would a simple cat command work?

Also, is this a silly thing to do? (I'm a newbie) Ultimately my goal is to determine gene expression changes.

Cheers.

[arkal]

[QUOTE=krespim;80152]

I've downloaded all the files, however it seems that each replicate/experiment has several file corresponding to several runs. The question is: how to concatenate these runs once they are converted into fastq? Would a simple cat command work?
[./QUOTE]

I'm sorry, why exactly do you want to cat all these files? A simple a simple
cat *.fastq >> consolidated_fastq.fq

will be fine but whats your need for doing so? Every single run usually corresponds to a different sample so why merge all?

Also, is this a silly thing to do? (I'm a newbie) Ultimately my goal is to determine gene expression changes.

Cheers.

Process files separately then use comparitive studies on the sam/bam files!

[krespim]

will be fine but whats your need for doing so? Every single run usually corresponds to a different sample so why merge all?

Well, this is actually my main issue as I don't know if each run is a different sample, or the same sample ran in multiple lanes. The sample GEO page lists 3 SRA files.

The paper does not mention biological or technical replicates.

[arkal]

Well, this is actually my main issue as I don't know if each run is a different sample, or the same sample ran in multiple lanes. The sample GEO page lists 3 SRA files.

The paper does not mention biological or technical replicates.

It seems to be the same sample in different lanes/runs... so you can either merge the fastqs and align or align the 3 separately and merge the sams! I recommend the latter as it will take less time (provide you have the resources to align them parallelly)!

[krespim]

It seems to be the same sample in different lanes/runs... so you can either merge the fastqs and align or align the 3 separately and merge the sams! I recommend the latter as it will take less time (provide you have the resources to align them parallelly)!

I have recently been granted access to a server with a very decent number or processors so that should be easy

Thanks a lot arkal!

[arkal]

No worries all the best!