Add error messages to MAE shell scripts #175
Conversation
There was a problem hiding this comment.
The commits are fine just expend the error message for the user.
I just have one issue to raise here, but do not have the best solution yet. If you throw an error, the user can not finish the pipeline. Hence, he will not get the nice HTML reports before fixing those issues nor get a result file as it depends on all input of all samples. If you are happy with this for the end-user go ahead.
Maybe @vyepez88 has more throughs on this.
| num_out=$(zcat "${output}" | wc -l ) | ||
| if [ "${num_out}" -lt 2 ] | ||
| then | ||
| echo 'ERROR: No allele-specific counts' |
There was a problem hiding this comment.
Please add a suggestion how to solve it.
- Checking the chromosomes (bam/vcf)
- Check the reference -> having 1 reference/fasta file per BAM could also solve dome of the issues here
- Removing the sample from MAE
- ...
There was a problem hiding this comment.
- This shouldn't happen with the new fixes
- Nick is implementing this feature
| echo "ERROR: no entries after filtering for SNVs" | ||
| echo "VCF ID: ${vcf_id}" | ||
| echo "VCF file: ${vcf_file}" | ||
| echo "BAM file: ${bam_file}" | ||
| exit 1 |
There was a problem hiding this comment.
Same as above. Please add suggestions for the user to it.
- Do we do not have any heterozygous variants in the VCF file?
- Is the VCF file not correctly formated?
- Do we have issues with the chromosome style? (Not sure if this is really a problem in this script)
- ....
There was a problem hiding this comment.
number 3. shouldn't be a problem in this script.
4. Check if vcf file exists and its index exists (although the pipeline might not even begin if this is not the case)
|
I'm fine with the users not getting the aggregated results table + html. They can anyway get the results of all the other samples by running the pipeline with the |
|
@c-mertes Would you suggest to change this error message to a warning then and deal with empty ouputs later in the pipeline? |
Major changes included in this update is: * adapt to new snakemake version (#208) * Speedup CI (use micromamba, #218) * Add HTML cutoff for results (#217) * Separate modules (#209) * Update documentation (#228) On top of this many small bug-fixes are included: * #175 #192 #198 #231 * Simplify result reporting Co-authored-by: Vicente Yepez <30469316+vyepez88@users.noreply.github.com> Co-authored-by: Vicente <yepez@in.tum.de> Co-authored-by: Michaela Mueller <51025211+mumichae@users.noreply.github.com> Co-authored-by: Michaela Mueller <mumichae@in.tum.de> Co-authored-by: Smith Nicholas <smith@in.tum.de> Co-authored-by: nickhsmith <smithnickh@gmail.com> Co-authored-by: Christian Mertes <mertes@in.tum.de>
* Update to version 1.1.0 (#229) Major changes included in this update is: * adapt to new snakemake version (#208) * Speedup CI (use micromamba, #218) * Add HTML cutoff for results (#217) * Separate modules (#209) * Update documentation (#228) On top of this many small bug-fixes are included: * #175 #192 #198 #231 * Simplify result reporting Co-authored-by: Vicente Yepez <30469316+vyepez88@users.noreply.github.com> Co-authored-by: Vicente <yepez@in.tum.de> Co-authored-by: Michaela Mueller <51025211+mumichae@users.noreply.github.com> Co-authored-by: Michaela Mueller <mumichae@in.tum.de> Co-authored-by: Smith Nicholas <smith@in.tum.de> Co-authored-by: nickhsmith <smithnickh@gmail.com> Co-authored-by: Christian Mertes <mertes@in.tum.de> * add splitting to MAE filterSNVs.sh * fix broken link and maxVarFreqCohort * fix broken link and maxVarFreqCohort * Update test_MAE.py * code review fixes Co-authored-by: Michaela Mueller <51025211+mumichae@users.noreply.github.com> Co-authored-by: Vicente Yepez <30469316+vyepez88@users.noreply.github.com> Co-authored-by: Vicente <yepez@in.tum.de> Co-authored-by: Michaela Mueller <mumichae@in.tum.de> Co-authored-by: Smith Nicholas <smith@in.tum.de> Co-authored-by: Christian Mertes <mertes@in.tum.de>
Whenever the output of GATK ASECounter is empty, DROP will continue with the empty files until DESeq2 throws an error.
One option is to throw an error every time the ASECounter output is empty. This is implemented here. There might be other possibilities in addressing this issue, so any discussion is welcome.