How to processing ENCODE small RNA-seq data

[dzmtnvmt]

Here is a problem when processing CshlShortRnaSeq data from ENCODE.

For example, for 1*36bp Gm12878, I found the adaptor or primer sequence of small RNA library were:
-------------------------------------------------------------------
5’SBS3_Adapter (This is the RNA ligated onto the 5’ end): “r” = ribose, RNA base
5’- rArCrArCrUrCrUrUrUrCrCrCrUrArCrArCrGrArCrGrCrUrCrUrUrCrCrGrArUrCrU
A-Tail RT Primer (This is the primer used in the RT reaction):
5’-TCTCGGCATTCCTGCTGAACCGCTCTTCCGATCTTTTTTTTTTTTVN
PE 5’ PCR (PCR Primer):
5’-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATC
PE 3’ PCR (PCR Primer):
5’-CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTC
----------------------------------------------------------------------

but I found that there were few reads with the adaptor on the 5' side, but rather, there are many "AGATCGGTTGT*" (the reverse of 5'adaptor) after the ployA in the 3' side. So, I am not sure if it would be correct if I clip the 5'SBS3_adaptor and 3'ploy A, and process those data accuratly using aligners such as Bowtie.

I will be appreciative if anyone could help~

[alexdobin]

At the 3' ends of your reads you should see the reverse complementary to the A-Tail RT Primer sequence: 5’-TCTCGGCATTCCTGCTGAACCGCTCTTCCGATCTTTTTTTTTTTTVN
i.e.
AAAAAAAAAAAAGATCGGAAGAGCGGTTCAGCAGGAATGCCGAGA

We trimmed any sequence that contained AAAAAA (6As). This is a very aggressive trimming - we hoped to get rid of all the genomic A-homopolymer priming sites. STAR can do it for you with:
--clip3pAdapterSeq AAAAAA --clip3pAdapterMMp 0

[dzmtnvmt]

At the 3' ends of your reads you should see the reverse complementary to the A-Tail RT Primer sequence: 5’-TCTCGGCATTCCTGCTGAACCGCTCTTCCGATCTTTTTTTTTTTTVN
i.e.
AAAAAAAAAAAAGATCGGAAGAGCGGTTCAGCAGGAATGCCGAGA

We trimmed any sequence that contained AAAAAA (6As). This is a very aggressive trimming - we hoped to get rid of all the genomic A-homopolymer priming sites. STAR can do it for you with:
--clip3pAdapterSeq AAAAAA --clip3pAdapterMMp 0

Thanks for your reply. Do you imply that the 5'SBS3_adaptor has already been trimmed in the raw fastq data?
BTY, I found that the 5’SBS3_Adapter and A-tail primer all have "CGCTCTTCCGATCT". Is there any meaning from this?

[alexdobin]

Under normal conditions, the 5' adapter does not get sequenced - the sequencing starts from the 1st base of the RNA sequence. The only sequence you have to worry about is the 3' adapter.

[dzmtnvmt]

Got that, thank you!