...
Here we will assume you have data from GSAF's Illumina HiSeq or MiSeq sequencer.
Learning Objectives
...
The GSAF website describes the flavaors of Illumina adapter and barcode sequence in more detail https://wikisutexas.utexasatlassian.edunet/wiki/display/GSAF/Illumina+-+all+flavors
Cutadapt
The cutadapt program is an excellent tool for removing adapter contamination. The program is not available through TACC's module system but we've installed a copy in our $BI/bin directory.
...
Expand | |||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||
Please refer to https://wikisutexas.utexasatlassian.edunet/wiki/display/GSAF/Illumina+-+all+flavors for Illumina library adapter layout. The top strand, 5' to 3', of a read sequence looks like this.
The -a argument to cutadapt is documented as the "sequence of adapter that was ligated to the 3' end". So we care about the <Read 2 primer> for R1 reads, and the <Read 1 primer> for R2 reads. The "contaminent" for adapter trimming will be the <Read 2 primer> for R1 reads. There is only one Read 2 primer:
The "contaminant" for adapter trimming will be the <Read 1 primer> for R2 reads. However, there are three different Read 1 primers, depending on library construction:
Since R2 reads are the reverse complement of R1 reads, the R2 adapter contaminent will be the RC of the Read 1 primer used. For ChIP-seq libraries where reads come from both DNA strands, the TruSeq Read 1 primer is always used.
For RNAseq libraries, we use the small RNA sequencing primer as the Read 1 primer.
|
...
Trimmomatic offers similar options to Flexbar with the potential benefit that many illumina adaptor sequences are already "built-in". It is available here.
More Example Data
See if you can figure out what's wrong with these data sets (copy them to your $SCRATCH
directory before analyzing them) and then process them to get rid of the problem(s). If you're very ambitious, you could also map them to the reference genomes and perform variant calling before and after cleaning them up to see how the results change. Each file has a different problem.
Example #1: Single-end Illumina MiSeq data for E. coli
Code Block | ||||
---|---|---|---|---|
| ||||
$BI/gva_course/read_processing/JJM104_TAAGGCGA-TAGATCGC_L001_R1_001.fastq.gz
$BI/gva_course/read_processing/REL606.fna |
Expand | ||
---|---|---|
| ||
This |
Example #2: Paired-end Illumina Genome Analyzer IIx data for E. coli
Code Block | ||||
---|---|---|---|---|
| ||||
$BI/gva_course/read_processing/61FTVAAXX_2_R1_ZDB172.fastq.gz
$BI/gva_course/read_processing/61FTVAAXX_2_R2_ZDB172.fastq.gz
$BI/gva_course/read_processing/REL606.fna |
Expand | ||
---|---|---|
| ||
There was some sort of problem during library prep that highly biased the beginning of reads to "T". Unfortunately, post-processing can't help with this one. The read sequences are fine, but the coverage across the genome is so uneven that many regions of the genome were not sampled (have zero coverage) even though the volume of sequencing data was very high for this microbial genome. The facility had to do a new library prep and re-sequence to correct this issue. |