cutadapt questions on pair-end seqeuencing data

[bbm]

I'm trying to use cutadapt to trim my data. The adaptor sequences from Illumina only has a universal adapter, then one sequence for each index.
Example here:
TruSeq Universal Adapter
5’ AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT
TruSeq Adapter, Index 1 5
5’ GATCGGAAGAGCACACGTCTGAACTCCAGTCACATCACGATCTCGTATGCCGTCTTCTGCTTG
TruSeq Adapter, Index 2
5’ GATCGGAAGAGCACACGTCTGAACTCCAGTCACCGATGTATCTCGTATGCCGTCTTCTGCTTG
TruSeq Adapter, Index 3
5’ GATCGGAAGAGCACACGTCTGAACTCCAGTCACTTAGGCATCTCGTATGCCGTCTTCTGCTTG

When I read the guide from cutadapt for pair-end data, they suggest a file named ADAPTER_FWD (should be trimmed from the forward reads (filename.1.fastq.gz) and a file named ADAPTER_REV from the reverse reads (filename.2.fastq.gz).
So my question is: What sequence do I put in ADAPTER_FWD, and what sequence in ADAPTER_REV?


The command is:
cutadapt -a ADAPTER_FWD -A ADAPTER_REV -o out.1.fastq -p out.2.fastq reads.1.fastq reads.2.fastq

[bbm]

Another related question here:
My sequencing center said they didn't trimmed any adapters. So I went to the fastq.gz files, and grep the adapter sequence for the first 1 million reads, I only have 10442 hits. I wonder why the hits number is so small? Does that mean 989,558 reads are shorter and without the adapters?
Thanks!

[Brian Bushnell]

Most sequences will be shorter than full length, and/or contain errors, and thus not match exactly with grep.

I recommend you use BBDuk for adapter trimming, using the command at the top of the post.