fastp
fastp copied to clipboard
fastp needs a directory to write output sequences
Minor thing that I've found through trial and error for fastp-v0.19.5 & fastp-v0.20.1 I cannot install the latest version of fastp; different issue: https://github.com/OpenGene/fastp/issues/383#issuecomment-1088442268
Regardless, here is the behavior I've noticed: fastp requires a directory to exist before you can write the output file in a different directory.
For example, given a directory of samples where the directory "fastq" is the parent directory, and each sample directory contains paired reads, and lastly the intended target directory is currently empty output directory (trimmed/)
data/
|__ fastq/
| |
| |__ sample_SRS111/
| | SRR119read1.fq.gz
| | SRR119read2.fq.gz
| |
| |__ sample_SRS112/
| SRS219read1.fq.gz
| SRS219read2.fq.gz
|
|__ trimmed/
|
|__ bash-scripts/
fastp_trim.sh
and given the basic fastp command:
fastp
-i data/fastq/sample_SRS111/SRR119read1.fq.gz
-o data/trimmed/sample_SRR111/SRR119_r1.fq.gz
-I data/fastq/sample_SRS111/SRR119read2.fq.gz
-O data/trimmed/sample_SRR111/SRR119_r2.fq.gz
fastp will not write the file unless there is a directory available for the output of fastp. It expects the directory to look like this:
data/
|
|__ fastq/
| |
| |__ sample_SRS111/
| | SRR119read1.fq.gz
| | SRR119read2.fq.gz
| |
| |__ sample_SRS112/
| SRS219read1.fq.gz
| SRS219read2.fq.gz
|
|__ trimmed/
| |__ sample_SRS111/
| |
| |__ sample_SRS112/
|
|__ bash-scripts/
fastp_trim.sh
This behavior may be intended, but I did not see it documented.