Small RNA adapters skip first 5 bases of R1?

[pmiguel]

I've now seen a couple of sets of libraries with Illumina "small RNA" adapters that seem to somehow "skip" the first 5 bases of R1. This being being obvious only when comparing the forward and reverse reads of the libraries. The apparent insert lengths are 5 bases longer in the R2 reads than the R1 reads. And that difference appears to be the lack of the very first 5 bases of the R1!?!

Has anyone else run small RNA libraries on a paired end run and checked for such a difference between the forward and reverse reads?


------
More details:


We first noticed this on the MiSeq, but looking back we also saw this issue on an earlier set of libraries that were run on both the MiSeq and on the NovaSeq.

Although I don't think it is the issue, both sets of libraries are odd--in different ways. The recent set is just 3 libraries--not made from small RNAs at all, but their adapters are those of canonical Illumina small RNA libraries. With adapters:

Oligonucleotide sequences©2017 Illumina,Inc. All rights reserved.

AATGATACGGCGACCACCGA CAGGTTCAGAGTTCTACAGTCCGA

CAAGCAGAAGACGGCATACGAGAT NNNNNN GTGACTGGAGTTCCTTGGCACCCGAGAATTCCA

with NNNNNN being the i7 index.

Oligonucleotide sequences©2017Illumina,Inc. All rightsreserved. Derivative works created by
Illumina customers are authorized for use with Illumina instruments and products only. All other
uses are strictly prohibited.

except that the left (i5) adapter has an extra 5' C to create an MmeI site.

And the earlier libraries were more normal small RNA libraries made with a Bioo (PerkinElmer) kit, but the left (i5) adapter is modified to contain an index:

AATGATACGGCGACCACCGAGATCTACACNNNNNNNNACCGAGATCTACACGTTCAGAGTTCTACAGTCCGA

so as to allow us to make UDI small RNA libraries.

--
Phillip

[pmiguel]

The answer is:

The sequence:
Oligonucleotide sequences©2017 Illumina,Inc. All rights reserved.

AATGATACGGCGACCACCGA CAGGTTCAGAGTTCTACAGTCCGA

is just the PCR primer and doesn't contain the final 5 bases of the adapter. The full adapter sequence would be:

AATGATACGGCGACCACCGA CAGGTTCAGAGTTCTACAGTCCGA CGATC

(spaces added for clarity...)

When I use the full adapter, the 5 base descrepancy disappears!

(Thanks to Illumina Tech support for finding this error in my thinking.)

--
Phillip