reversing a read of a paired-end

[lre1234]

Currently my adviser is interesting in working with single-ended RNA-seq datasets in order to maintain the strandedness information. To increase the number of sample that we can look at, we were thinking downloading some public datasets. As there are many PE RNA-seq samples available, my advisor suggested that we simply reverse the sequence read of one of the pairs in the PE, which would then maintain the stranded information. My thought is that this would not work and even if one read is flipped, then we are still not maintaining the strand information, but I might be wrong about this? As an alternative approach, we could just work with one of the reads in the pair, but we loose all of the reads from the other half. (My thought is that we should just work with the paired-end and look for splicing and gene-fusions which is more interesting than the strandedness information, but I seem to be fighting a loosing battle in this).

Any thoughts?

[JackieBadger]

When you say "maintain stand information" do you mean you just want PE R1 and R2 to be in the same orientation i.e both 5'->3' ?
If so you just reverse complement R2 (at least for Illumina data)

Most aligners/PE contig builders are smart enough to do this for you, but if you wanted to do it independently I think FASTX toolkit has a reverse complement function.

If your PE reads have some overlap, just use FLASH to construct contigs.

[mastal]

reversing a read of a paired-end

you can only get strandedness information from an RNA-Seq experiment if the sample is prepared using a strand-specific protocol.

If you are going to download publically available datasets, look for ones that have been done using a strand-specific protocol.

Otherwise, at least in the Illumina system, you get read1 from one strand, and read2 from the opposite strand, but this tells you nothing about which strand the original RNA transcript came from.

[lre1234]

Otherwise, at least in the Illumina system, you get read1 from one strand, and read2 from the opposite strand, but this tells you nothing about which strand the original RNA transcript came from.

This was my thought that 1 read was from 1 strand and 1 read was from the opposite strand. I didn't realize that it could be done using a strand-specific protocol. I guess I have to look for those!

Thanks for the help!

[GenoMax]

Here is a specific example of stranded RNA-seq: http://www.ncbi.nlm.nih.gov/geo/quer...i?acc=GSE44450

Searching for "stranded RNAseq" on SRA may help but you will need to sift through the metadata to know for sure if a stranded method was used.