How to work with ENCODE RNA-seq fastq files?

[Karenj]

I have downloaded the fastq files for paired end RNA-sequencing for a single cell line from the ENCODE project, as I want to map the reads my self in the search for chimeric read pairs.
My problem is that for each sample, the read1 and read 2 fastq.tgz folders contain 5 separate fastq-files. Should these files be merged before or midway in my mapping? If yes - how?

[alexdobin]

Hi @Karenj,

which particular files did you download and where did you download them?
My guess is you are seeing the files separate by Illumina lanes.
The lanes are supposed to be merged on the UCSC server for most of the samples. If not, you can merge them before or while mapping. It's important that the order of merging is exactly the same for both reads 1 and 2.

Alex

[Karenj]

Hi Alex,

The files are RNA-seq for K562 from the Caltech data set, 2x75 paired end. I found few minutes ago, that for the CSHL RNA-seq data set the fastq files are merged and the data set is newer, so perhaps it is easier to work with this.

[alexdobin]

I would also recommend working with CSHL dataset
It's indeed newer and is more uniform in terms of quality and sequencing depth.
Note, that CSHL's dataset is stranded dUTP protocol, while Caltech's is mostly standard unstranded.

[Karenj]

Thanks for your answer. I will try to see if the data from CSHL behaves better :-)