Illumina HiSeq2500 Pipeline Question

[hamcan]

Hi there,
I have just received my Illumina HiSeq data back. I have 2x250bp runs (so forward and reverse reads).

I ran a FASTQC and found that I have an adaptor sequence present in my data. Should I first remove the adaptors then merge my forward and reverse reads using FLASh? Or should I merge first, then remove the adaptors and trim at the same time (using trimmomatic).
Thank you!

[mastal]

Remove the adapters first.

Note that for the reads that have adapters, the fragment length is less than 250 bp, and your forward and reverse reads, after adapter removal, should be reverse complements of each other.

[hamcan]

Remove the adapters first.

Note that for the reads that have adapters, the fragment length is less than 250 bp, and your forward and reverse reads, after adapter removal, should be reverse complements of each other.

So should I first use trimmomatic then to remove the adaptors and to overall trim the sequences at once? Then merge?

Thank you!

[Zapages]

Merging the reads into one file depends on the task that you are working on.

and also do some quality control on the sequences.

Please FastQC to check the quality of after each trimming and/or quality control trimming.

[hamcan]

Merging the reads into one file depends on the task that you are working on.

and also do some quality control on the sequences.

Please FastQC to check the quality of after each trimming and/or quality control trimming.

Okay thanks!!
I am planning on doing cutadapt to remove the adaptor sequences. Then I will merge my sequences and then Trim them using trimmomatic so that I can blast them against the NCBI-nr database. This is an environmental metagenomics sample and I'm looking to see what bacterial species are present in certain environments.

[lwebs]

I used FLASH v1.2.11 to merge forward and reverse reads from a HiSeq4000 run on environmental DNA (estimated frag length 280-700 bp, read length 150 nt) and am obtaining really low read overlap ~18.40% (counting both innie and outie pairs). Is this to be expected? Any help/insight would be appreciated!

[GenoMax]

I used FLASH v1.2.11 to merge forward and reverse reads from a HiSeq4000 run on environmental DNA (estimated frag length 280-700 bp, read length 150 nt) and am obtaining really low read overlap ~18.40% (counting both innie and outie pairs). Is this to be expected? Any help/insight would be appreciated!

Looking at your fragment lengths it sounds reasonable. Once your fragment size goes over ~ 300 bp R1/R2 will no longer overlap at 150 bp length.

You could try BBMerge from bbmap suite to see if that does any better in terms of getting better merging.