What is the relationship between Illumina P5 P7, universal adapters, and TruSeq Adapter Index 7?

Question

[octpus616]

Hi,

I am processing a batch of sequencing data, which comes from special library construction technology, and it may contain more adapter sequences than standard library data. I have been informed that the adapter sequences used in library construction are:

> P7_adapter
AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC
> P5_adapter
AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT

First, I tried to directly remove these sequences using cutadapt with params -a AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC -A AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT, but it was not effective. FastQC shows that there are still a lot of Illumina universal adapters in R2.

Then I tried using shorter adapter sequences -a AGATCGGAAGAG -A AGATCGGAAGAG, and in this case, it worked very well. However, my fastqc report shows that TruSeq Adapter, Index 7 sequences are overrepresented. I'm not clear about the reason for this phenomenon and would like guidance on how to remove these sequences.

Answer 1

[GenoMax]

Illumina "universal" adapter is designated as AGATCGGAAGAG in FastQC adapter configuration file. You can see that it is basically the beginning of the P7 and P5 adapters you note above.

You may need to tell us more about the "special library" construction. It is possible that this could be a red herring and FastQC is simply picking on some similarity there and it can be ignored. Once the P7/P5 adapter start is found all sequence to the 3-end side should be eliminated by trimming programs.

[octpus616]

Hi, GenoMax

Thanks for you kindly help, basicly, we follow a enhance clip protocol (https://doi.org/10.1038/nmeth.3810), since the template has been digested, the fragment length may be shorter, so we think it is reasonable to find the adapter sequence at the 3' end and it has been removed by cutadapt. However, I am not sure whether the index sequence detected by FastQC is a false alarm, really exists, or can be ignored.