recommendation-for-processing-forward-and-reverse-reads-with-low-quality-issues

[tuyick (Patrick)]

Hello

I would like to request some recommendations-for-processing-forward-and-reverse-reads-with-low-quality-issues. one person told me that once I have such problem may be I can process the forward reads only. any one to help me to find some code that I can use to proceed with with forward reads only while I'm using Casava paired-end demultiplexed fastq files

[szymanski (Shay)]

If you are using Casava Paired-end demultiplexed fastq files, there should be a general format that names are in that will indicate which read they belong to. For example, in an experiment I did, I would have file names like 5-12-21-BM-4_R1_001.fastq.gz and 5-12-21-BM-4_R2_001.fastq.gz, in which R1 was my forward reads and R2 is my reverse reads. So, I have a directory that looks like this:

5-12-21-BM-1_R1_001.fastq.gz
5-12-21-BM-1_R2_001.fastq.gz
5-12-21-BM-2_R1_001.fastq.gz
5-12-21-BM-2_R2_001.fastq.gz
5-12-21-BM-3_R1_001.fastq.gz
5-12-21-BM-3_R2_001.fastq.gz
5-12-21-BM-4_R1_001.fastq.gz
5-12-21-BM-4_R2_001.fastq.gz

If I wanted to keep just my forward reads, by command line I could do something like
rm *_R2_001.fastq.gz
which would remove (rm) my files that end with _R1_001.fastq.gz
Depending on if you are using Linux, Mac, or Windows, the specific command may vary. From here you could import the remainder. If you want to be safer, you could instead make a subdirectory for R1 and R2, and move each file with a certain format into the subdirectories

mkdir R1
mkdir R2
mv *_R1_001.fastq.gz R1
mv *_R2_001.fastq.gz R2
Again, commands may vary based on your system.

More information would be helpful to get more help!
Big ones that I have are:
What is your target region (16S v3/v4 subunit, ITS1, ITS2, 18S, etc.)?
Primer pair? (very relevant for ITS in my experience)
What is your sequencing platform specifications (things like 150bp Paired end or 300bp paired end, illumina miseq vs nanopore, etc).
What do you mean by low quality exactly (low reads/sample? poor quality scores throughout the sample? quality score drop at end?)?

Depending on the answers to these questions, the processing you can do and usefulness of things like read joining will vary.