Hi, It's me again. I have issue with the following error:

ERROR ~ Failed to invoke `workflow.onComplete` event handler

and was told to refer to line 195 in main/nf in which I did. It was about intervals and considering that my sequences are whole genome sequencing, therefore there is no need for interval analysis (if my understanding is correct). The following is the params file that I have tweaked into.

{
    "input": "\/rds\/general\/user\/ksheng\/projects\/molecularembryologydata\/live\/Samantha\/Low_Pass_Seq_Batch_2\/sample_sheet.csv",
    "outdir": "\/rds\/general\/user\/ksheng\/projects\/molecularembryologydata\/live\/Samantha\/Low_Pass_Seq_Batch_2\/Test_13",
    "intervals": false,    
    "tools": "controlfreec",
    "trim_fastq": true,
    "save_mapped": true,
    "save_output_as_bam": true,
}

My command used to run Nextflow:

nextflow run nf-core/sarek -r 3.4.1 -name "Test_37" -c /rds/general/user/ksheng/home/config.modified -work-dir /rds/general/user/ksheng/ephemeral/SAREK/work  -resume -params-file "nf-params.json"

Despite this, I am still met with the same error. No output for variant calling. My initial problem was that at the end of each run, it yielded only results for 1 sample for variant calling. I have a total of 20 samples submitted for the analysis. To obtain the rest of the samples, I re-ran the same pipeline. However, the generated variant calling result is random, therefore I have no control over which sample will have the variant calling result. By repeating 20 times, I managed to obtain 6 samples, and many times it is the repetition of the same sample. Can anyone help me with this, please? Thank you.

The intervals file for WGS is the one from GATK: https://gatk.broadinstitute.org/hc/en-us/articles/360035531852-Intervals-and-interval-lists https://gatk.broadinstitute.org/hc/en-us/articles/360035889551-When-should-I-restrict-my-analysis-to-specific-intervals:

It is not strictly necessary to restrict analysis to intervals when working with whole genomes, since presumably you're interested in all of it. However, from a technical perspective, you may want to mask out certain contigs (e.g. chrY or non-chromosome contigs) or regions (e.g. centromere) where you know the data is not reliable or is very messy, causing excessive slowdowns. In addition, defining whole-genome intervals allows you to parallelize execution across intervals using the scatter gather mode of parallelism.

We share the lists of "good" whole-genome intervals that we use in our production pipelines for human analysis in our resource bundle (see Download page).

it depends on what you want to do with it whether or not this makes sense. But to ignore intervals altogether, you need to set

--no_intervals

otherwise the pipeline uses the fasta index to construct intervals (so it will use all contigs and paralleleises along them, while with the default intervals it will also exclude certain regions as defined by GATK). Can you send you entire

nextflow.log

file please?'

Hi @Rike Hanssen. Stupid question, how can I obtain nextflow.log file?

It should be in the directory where you executed the run. It starts with a

.

so full name is

.nextflow.log

I hope I got it right because there's no file starting with a

.

and ends with

.nextflow.log,

but I will show you based on what I think it should be.

can you do a

ls -lha

in the directory you ran this?

@Rike Hanssen I found it!

FYI, any file starting with

.

is "hidden" on Unixen, so you have to use

-a

with

ls

to show it and globs (like

*

) will ignore it by default, but you can do any other operation by referring to it with its full name.

Thank you for the heads up. 🙂

The

workflow.onComplete

error is caused by failure to find the

mail

program (

Caused by: java.io.IOException: Cannot run program "mail": error=2, No such file or directory

) to send an e-mail upon completion, which it seems you have set up. If you need an e-mail, you'll need to install

mail

and put it in your

PATH

, otherwise change the config to not send mail. The issue with the collapsed samples might be a name collision causing samples to overwrite each other. Seeing your sample sheet would be helpful in assessing this.

Hi @Thomas Danhorn, Thank you for the informative reply. I will remove the email path as I do not need it. I have attached the sample sheet used here. Please have a look at it. Thank you.

This looks good to me, all the samples are unique, so I don't know what is going on. If you look in the work directory, can you confirm that all of the samples were processed, but just not published in the results directory? Or were they not even processed?

@Thomas Danhorn Thank you for the feedback. Regarding checking the samples in the work directory, I scanned through them once but they were all written in numbers, I'm unsure how can I interpret them whether which is for which samples. Can you walk me through a little on how I can do this check?

See what

ls -l /rds/general/user/ksheng/ephemeral/SAREK/work/??/*.Aligned.out.bam | less

gives you -- one sample, or all of them? (Note that this may take several seconds to find the files. Press

q

to leave

less

.)