Deepvariant variant calling by singularity

Question

[fra.r.silvestro]

The code I run:

1. 
BIN_VERSION="1.5.0"

sudo apt -y update
sudo apt-get -y install docker.io
sudo docker pull google/deepvariant:"${BIN_VERSION}"

2.
# Pull the image
singularity pull docker://google/deepvariant:"${BIN_VERSION}"

3. 
mkdir 04.deepvariant 
# Run singularity
singularity run -B /mountpoint/fastQ/:/input \
  deepvariant_1.5.0.sif \
  /opt/deepvariant/bin/run_deepvariant \
  --model_type WES \
  --ref $REF  \
  --reads 03.align/$FQ.align.sort.marked.bam \
  --output_vcf /input/04.deepvariant/$FQ.output.vcf.gz

It doesn't work. The message:

INFO: converting sif file to temporary sandbox. INFO: cleaning up image.

It prints all the flags related to /opt/deepvariant/bin/run_deepvariant.py Maybe some flags are wrong. I don't know if the output directory could be the root directory where my reads and reference are located.

[fra.r.silvestro]

I moved my data (reference .fasta, .fai and .bam, .bai) inside the INPUT_DIR. Then edit the code following this instruction https://github.com/google/deepvariant/blob/r1.5/docs/deepvariant-quick-start.md:

PWD=/mountpoint/fastQ
INPUT_DIR="${PWD}/testdata_input"
mkdir -p ${INPUT_DIR}
OUTPUT_DIR="${PWD}/04.deepvariant_output"
mkdir -p "${OUTPUT_DIR}"

sudo docker run \
  -v "${INPUT_DIR}":"/input" \
  -v "${OUTPUT_DIR}":"/output" \
  google/deepvariant:"${BIN_VERSION}" \
  /opt/deepvariant/bin/run_deepvariant \
  --model_type=WES \
  --ref=/input/Homo_sapiens_assembly38.fasta \
  --reads=/input/$FQ.align.sort.marked.bam \
  --output_vcf=/output/$FQ.vcf.gz \
  --output_gvcf=/output/$FQ.g.vcf.gz

This the error message:

I0703 11:40:58.754097 140133281920832 make_examples_core.py:257] Overhead for preparing inputs: 6 seconds
I0703 11:40:58.768027 140133281920832 make_examples_core.py:257] 0 candidates (0 examples) [0.01s elapsed]

[amy__]

Have you tried to build the container first and got the .sif file?

[fra.r.silvestro]

Now I'm trying. By running:

singularity build --sandbox /mountpoint/fastQ/singularity_sandbox docker://google/deepvariant:"${BIN_VERSION}"

I got this ERROR:

Could not remove bundle: could not remove "/mountpoint/fastQ/build-temp-709555147": unlinkat /mountpoint/fastQ/build-temp-709555147/rootfs/opt/licenses.zip: operation not permitted

I don't think is the solution. The image pull (the .sif file) from the Docker hub should be right.

[fra.r.silvestro]

What do you fin for build the container? As the command have I posted? Otherwise, do you suggest a tutorial?

[amy__]

Hi, sorry for late reply - what are you building this on is it a HPC?

When I built my sif container I did it in the login node of my HPC and loaded singularity and then ran singularity pull docker://google/deepvariant:1.3.0

[fra.r.silvestro]

No HPC, just my personal computer to test it. https://github.com/google/deepvariant/blob/r1.5/docs/deepvariant-quick-start.md This is the tutorial I've followed. I'm using conda package manager to install singularity I get a .sif file.

I get this output:

I0710 10:57:37.057581 140272616249152 make_examples_core.py:257] Overhead for preparing inputs: 18 seconds
I0710 10:57:37.079038 140272616249152 make_examples_core.py:257] 0 candidates (0 examples) [0.02s elapsed] 

All flags are right, I hope.

[Philipp Bayer]

Your output should be in /mountpoint/fastQ/04.deepvariant/ outside of the Singularity image, is it not there?

are $REF and $FQ set to something? what does echo $REF and echo $FQ say?

[fra.r.silvestro]

$FQ is just the ID sample. $REF is the path of my reference .fasta file. I already edited --output_vcf with output directory as you suggested, but it doesn't work as well.

singularity run -B /mountpoint/fastQ/:/input \
deepvariant_1.5.0.sif \
/opt/deepvariant/bin/run_deepvariant \
  --model_type=WES \
  --ref=/input/references/Homo_sapiens/GATK/GRCh38/Sequence/WholeGenomeFasta/Homo_sapiens_assembly38.fasta  \
  --reads=/input/03.align/$FQ.align.sort.marked.bam \
  --output_vcf=/mountpoint/fastQ/04.deepvariant/$FQ.output.vcf.gz

[amy__]

Have you tried using the full file name to see if that works? Feel free to email me [e-mail]

[jeffhsu3]

I am having a similar issue was this ever resolved? Getting this error: I0922 22:22:08.794076 140356687390528 make_examples_core.py:257] Overhead for preparing inputs: 0 seconds I0922 22:22:08.795082 140356687390528 make_examples_core.py:257] 0 candidates (0 examples) [0.00s elapsed]

I'm pretty sure all the inputs are correct as I've bound each one individually, but the error still persists.