skipped alignment steps

[mlorthiois]

Hi,

When I use Nanoseq with .fasta5 files as input, all the alignement steps are skipped. Here are my commands:

nextflow run nf-core/nanoseq \
    -r 1.0.0 \
    -profile singularity \
    --input /home/genouest/cnrs_umr6290/mlorthiois/nanoseq_test/input.csv \
    --protocol cDNA \
    --input_path /home/genouest/cnrs_umr6290/mlorthiois/nanoseq_test/data \
    --flowcell FLO-MIN106 \
    --kit SQK-DCS109 \
    --skip_demultiplexing \
    --max_cpus 30 \
    --igenomes_ignore false 

At the end, I get -[nf-core/nanoseq] Pipeline completed successfully-, but only these steps are computed:

[86/2a19f8] process > CheckSampleSheet (input.csv)   [100%] 1 of 1 ✔
[5b/8e1c88] process > Guppy (data)                   [100%] 1 of 1 ✔
[42/1c9b63] process > PycoQC (sequencing_summary.... [100%] 1 of 1 ✔
[b5/a3702e] process > NanoPlotSummary (sequencing... [100%] 1 of 1 ✔
[-        ] process > NanoPlotFastQ                  -
[-        ] process > FastQC                         -
[-        ] process > GetChromSizes                  -
[-        ] process > GTFToBED                       -
[-        ] process > GraphMap2Index                 -
[-        ] process > GraphMap2Align                 -
[-        ] process > SortBAM                        -
[-        ] process > BAMToBedGraph                  -
[-        ] process > BedGraphToBigWig               -
[-        ] process > BAMToBed12                     -
[-        ] process > Bed12ToBigBed                  -
[23/a0803d] process > output_documentation           [100%] 1 of 1 ✔
[3f/323218] process > get_software_versions          [100%] 1 of 1 ✔
[99/fdd59c] process > MultiQC (1)                    [100%] 1 of 1 ✔

So I get a fastq.gz file by guppy, the only solution I found is to run again Nanoseq, without basecalling and demultiplexing, with the fastq.gz as input. In this case, all the steps are computed.

[drpatelh]

Hi @mlorthiois ! Apologies for the late response to finally attempt to fix this. Are you able to re-run the analysis using the the code on dev? We are planning to release this soon and if you are still having issues then we can try and fix it.

nextflow pull nf-core/nanoseq
<YOUR_COMMAND> -r dev

[mlorthiois]

Hi,
Thanks for the help, I'll run this quickly and come back to tell you if the problem is solved.

[mlorthiois]

Hi,

I re-run Nanoseq on the dev branch, with this command:

nextflow run nf-core/nanoseq \
    -profile singularity \
    --input input.csv \
    --protocol cDNA \
    --input_path /path/to/fast5_pass \
    --flowcell FLO-MIN106 \
    --kit SQK-DCS109 \
    --skip_demultiplexing \
    --max_cpus 30 \
    --igenomes_ignore false \
    --guppy_gpu true \
    -r dev

But I was not able to finish the analysis and see if all steps are completed, here is what I got before:

executor >  local (6)
[cd/582b09] process > CheckSampleSheet (input.csv)   [100%] 1 of 1 ✔
[62/8412d2] process > Guppy (fast5_pass)             [100%] 1 of 1 ✔
[f7/8133cd] process > PycoQC (sequencing_summary.... [100%] 1 of 1, failed: 1 ✘
[f1/36eefa] process > NanoPlotSummary (sequencing... [100%] 1 of 1 ✔
[-        ] process > NanoPlotFastQ                  -
[-        ] process > FastQC                         -
[-        ] process > GetChromSizes                  -
[-        ] process > GTFToBED                       -
[-        ] process > MiniMap2Index                  -
[-        ] process > MiniMap2Align                  -
[-        ] process > SortBAM                        -
[-        ] process > BAMToBedGraph                  -
[-        ] process > BedGraphToBigWig               -
[-        ] process > BAMToBed12                     -
[-        ] process > Bed12ToBigBed                  -
[-        ] process > Bambu                          -
[-        ] process > DESeq2                         -
[-        ] process > DEXseq                         -
[fb/3d017c] process > output_documentation           [100%] 1 of 1 ✔
[da/81e929] process > get_software_versions          [100%] 1 of 1, failed: 1 ✘
[-        ] process > MultiQC                        -
-[nf-core/nanoseq] Pipeline completed with errors-
WARN: To render the execution DAG in the required format it is required to install Graphviz -- See http://www.graphviz.org for more info.
Error executing process > 'PycoQC (sequencing_summary.txt)'

Caused by:
  Process `PycoQC (sequencing_summary.txt)` terminated with an error exit status (127)

Command executed:

  pycoQC -f sequencing_summary.txt -o pycoQC_output.html

Command exit status:
  127

Command output:
  (empty)

Command error:
  .command.sh: line 2: pycoQC: command not found

[drpatelh]

Thanks @mlorthiois. Ok, thats odd. The only reason I can think of is that we have had to use a separate container for pycoQC because I wasn't able to add it to the main Conda environment for the pipeline due to dependency conflicts.

That container I linked to above should get downloaded when you run the pipeline with the parameters you used though. Can you try and add export NXF_SINGULARITY_CACHEDIR=./containers/ before you run the command and it should download all of the containers for the pipeline in that directory.

Please also run nextflow pull nf-core/nanoseq beforehand too because I have merged some new changes into the dev branch to fix the other issues you were seeing.

[mlorthiois]

Thanks. I tried, but the new dev version tries to download the container nanoseq-1.1.0, without success.

Caused by:
  Failed to pull singularity image
  command: singularity pull  --name nfcore-nanoseq-1.1.0.img docker://nfcore/nanoseq:1.1.0 > /dev/null
  status : 255
  message:
    FATAL:   While making image from oci registry: error fetching image to cache: failed to get checksum for docker://nfcore/nanoseq:1.1.0: Error reading manifest 1.1.0 in docker.io/nfcore/nanoseq: manifest unknown: manifest unknown

So I'll wait until it is available on docker hub to retry.

[drpatelh]

Ok. Yes, sorry 🤦 This is turning into a bit of hack. Can you run with -r master instead. That should work.

[mlorthiois]

I have the same error ...

Caused by:
  Failed to pull singularity image
  command: singularity pull  --name nfcore-nanoseq-1.1.0.img docker://nfcore/nanoseq:1.1.0 > /dev/null
  status : 255
  message:
    FATAL:   While making image from oci registry: error fetching image to cache: failed to get checksum for docker://nfcore/nanoseq:1.1.0: Error reading manifest 1.1.0 in docker.io/nfcore/nanoseq: manifest unknown: manifest unknown

[mlorthiois]

Ok I've been able to launch the new version 1.1.0. I still have the same problem ...
I attached my stdout, my log file and here are my commands :

export NXF_SINGULARITY_CACHEDIR=./containers/

nextflow pull nf-core/nanoseq

nextflow run nf-core/nanoseq \
    -profile singularity \
    --input input.csv \
    --protocol cDNA \
    --input_path data \
    --flowcell FLO-MIN106 \
    --kit SQK-DCS109 \
    --skip_demultiplexing \
    --max_cpus 16 \
    --guppy_gpu true \
    -r 1.1.0

slurm-3327039.txt
nextflow.log

[drpatelh]

Can you delete your ~/.singularity directory and try again?

[drpatelh]

See here

[mlorthiois]

Done, but same problem ...

[drpatelh]

Ok. So now you don't have an issues with the download of the container? The pipeline finishes successfully but it still skips all of the alignment steps 🤦 Yeah, will need to keep this issue open then until I have time to fix this properly before the next release. Sorry, I thought the igenomes fix would work but it clearly hasn't...Thanks for bearing with me @mlorthiois 😎

[mlorthiois]

Oh yes, sorry, I didn't specify. Yes I downloaded the container without any problem. I just still have the initial problem of the skipped steps.
Thanks for the help, I'll keep looking on my side, if I find a solution I'll come back here 👍

[drpatelh]

This should be fixed when this commit is merged into dev @mlorthiois 34c7f08 . Will close this now but if the issue persists then please feel free to re-open. Apologies for the delay in getting around to fixing this and thanks for your help in debugging. Turned out to be a simple channel issue with the way the info is provided for data that requires basecalling but not demultiplexing.

Merry Xmas and Happy New Year!

[MartinezRuiz-Carlos]

Hello,

I am having the exact same issue under the same exact circumstances. I am skipping de-multiplexing and running guppy on a gpu.
It runs through, but all alignment steps are skipped, this issue is not solved when using -r dev
Here is the command I ran

nextflow run nf-core/nanoseq \
  --input samplesheet.csv \
  --protocol DNA \
  --input_path path/to/fast5/ \
  --flowcell FLO-PRO002 \
  --kit SQK-LSK110 \
  --skip_demultiplexing \
  --skip_quantification \
  --aligner 'minimap2' \
  --guppy_gpu \
  --gpu_cluster_options '--part=gpu --gres=gpu:1' \
  -profile crick \
  -resume