bbmerge detects different adapters

[Chief_Lazy_Bison]

Hello,

I have some bacterial genomes which were sequenced using the Nextera Flex library prep kit ( insert size ~350 bp) and MiSeq paired end reads (2x300). I want to merge these reads with bbmerge prior to adapter trimming and assembly.

I have used bbmerge's adapter detection feature but have gotten some strange results.
bbmerge seems to detect different adapters on each sample, and many have multiple N's.

The first part of the detected adapter sequence is what Illumina provides as the Nextera Flex adapter sequence, but I'm not sure what the other stuff is.

Here are some of the adapters that were detected:

>Read1_adapter
CTGTCNCTTATACACATCTCCGAGCCCACGAGACGGACTCCTANCTCGTATGCCGTCTTCTGCTTG

>Read2_adapter
CTGTCNCTTATACNCATCTGACGCTGCCGACGAAGAGGANAGNGNNNNNNNNGNNNGNNNC

This is the Nextera Flex adapter sequence given by Illumina CTGTCTCTTATACACATCT

as you can see the first part matches this except for an N stuck in.

Across my 20 samples 16 unique Read1 adapters were detected and
18 unique Read2 adapters were detected.

Should I worry about this? should I feed the detected adapter sequences into bbmerge? should I just give bbmerge the Illumina provided sequence?

Thanks!

[GenoMax]

I would suggest that you use the Illumina provided adapter sequence. BBMerge detection feature is good when you don't have that information a priori. There may be some sequencing errors in your reads which is leading to that N insertion.