kmer not removed by Trimmomatic

[Niranjanks]

I performed FastQC on paired end data and performed Trimmomatic trimming of TruSeq3-Pe-2.fa

After trimming, I see that the files fail for the kmer AGATCGG. 0 over-represented sequences were found.

However this kmer is already present in my Trimmomatic adapter!

>PE1_rc
AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTA

What should I do? Since I want to perform assembly next I am worried that further tools like BBDuk given in other threads wil reduce some DNA quality.

[mastal]

What parameters have you been using with ILLUMINACLIP?

The reason trimmomatic doesn't remove it is that even a perfect match for a 7-mer will score too low to be recognised as an adapter seq that should be trimmed.

Also, according to the Trimmomatic manual, the format for the ILLUMINACLIP command is

ILLUMINACLIP:<fastaWithAdaptersEtc>:<seed mismatches>:<palindrome clipthreshold>:<simple clip threshold><minAdapterLength>:<keepBothReads>

with the default value of minAdapterLength being 8.

Try setting a smaller value for minAdapterLength.

[Niranjanks]

ILLUMINACLIP:TruSeq3-PE-2.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:40

is what I used

[GenoMax]

What should I do? Since I want to perform assembly next I am worried that further tools like BBDuk given in other threads wil reduce some DNA quality.

That is not true. Any/all these programs are going to trim data based on parameters you use so on their own they will not reduce quality. In any case you want to eliminate extraneous sequence from your dataset if you want to assemble the data.

[Niranjanks]

Thank you for your replies. I will try it out.