MultiQC doesn't produce data for all samples, why? (trimmomatic)

Question

[john]

I've done a Paired-End trimmomatic run for 20 samples (40 if you count forward and reads as separate). When I run the

MultiQC ./

command, I'm not getting data for all 40 samples, rather I only get data from 19 samples and additionally, sample 20 is left out and with the exception for sample 1, I only get data on reverse reads. Any idea why that may be?

My code:

for file in ./*.fastq.gz; do trimmomatic PE -threads 1 -trimlog $file_trim.log -basein $file -phred33 -baseout ${file/%fastq.gz/fq.gz} ILLUMINACLIP:TruSeq3-PE.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:100 CROP:100 done

Answer 1

[Phil Ewels]

Take a look at the verbose log (-v or multiqc_data/multiqc.log) and have a read of https://multiqc.info/docs/#not-enough-samples-found

Most of the time it's either large log files or sample names being overwritten.

[john]

Thank you. I did take a look, and I've tried adding the -d and -s flags, but that doesn't help. All my files have unique names, so I don't think that's the problem?

[Phil Ewels]

Do you see anything in the verbose log about overwriting sample names? Note that it may be taking sample names from the input FastQ files and not the log file for example.

Also double check the log file sizes. The default limit is fairly low, maybe 10MB off the top of my head?

If neither of those help, please make an issue on the repo and attach the verbose log file. Ideally, try to reduce the problem as much as possible - for example down to just 2 or 3 samples which show the error.