Seqr installation on windows

[AhmadSafeer12345]

Helo Team,

I am trying to install Seqr locally on a windows system. I looked over the instructions and it seems to be only for linux. If you can guide about this, it would be grateful.

Thanks!

[hanars]

Hi Ahmad,

We do not have any guidance for installing seqr on a windows operating system, it is not a supported OS. If you are able to get it to work we would love to hear from you what modifications you needed to made! However, it is probably better for you to run it on a linux VM.

[AhmadSafeer12345]

Helo,
Thank you for your response! I have locally installed Seqr on Linux and created a new project, but I didn't see anywhere I can upload my VCF files and start the analysis. Could you please let me know how I can proceed with that?

[hanars]

You can find instructions for how to load data here: https://github.com/broadinstitute/seqr/blob/master/deploy/LOCAL_INSTALL.md#annotating-and-loading-vcf-callsets

[AhmadSafeer12345]

Hi @hanars , we have done all the steps and loaded the data in seqr, which you can see in the picture below:

Now we try to load this data into our project as described here: Adding dataset to seqr project
We have now made the project and try to load this data as described above but get the error which you can see in the screenshot below:

Now what could be the fix for this issue. Do we need to create the project in a specific way? Kindly respond to this as soon as possible as this is hindering my workflow.

[hanars]

You need to create families and individuals in the project, using the "edit families and individuals" button

[AhmadSafeer12345]

Helo, I have created families and individuals as shown in the given image.

Now, when I am trying to load data, it shows an error, as shown in the given image.

what should I do?

[AhmadSafeer12345]

Hi,
I used the exact same instructions. However my existing container was named "ahmasafe-seqr-1". When I did
docker-compose up -d
It started creating a new container named "seqr-seqr1" but failed to do so since the port was already allocated. My question is why did it create this new container when there was already one running there with my data? For this purpose now I had to stop the existing container "ahmasafe-seqr-1" and then again ran above command which created a new container but now since it was a new container and wasn't mapped to the existing volumes of the original one, hence it didn't contain any of the new data. Now how do I run the command with my existing container? This is a very weird kind of error.

[hanars]

I'm not really sure why your container was named ahmasafe in the first place - this sounds like you have changed your environment or where you were running these commands from in between runs. Essentially what you did was instead of updating your existing seqr deployment you created a different deployment in a different context. This is not something the seqr team can help you with, as making sure the environment and conditions under which you deploy seqr are maintained is your responsibility

[AhmadSafeer12345]

Yes, I have not managed to update the seqr version and its working fine now in the new version. I updated the paths for my custom usage in docker-compose and now able to get the old data as well. But just wanted to point out one thing. Whenever I run docker-compose up -d, it spins up the containers. However, for elastic search container it fails to run initially since it doesn't have the required permissions and I have to run the below command each time before to give it permission to access local directory
`
sudo chown -R 1000:1000 ./data/elasticsearch

`

my question is why do I have to give this permission whenever doing docker-compose up. Secondly is there a way to persist this? If not, I would suggest to add this to the documentation.

[hanars]

I think this issue is probably related to the permissions you have on your machine and those directories, not an issue within the seqr docker compose, but thats not something I can verify. seqr is an open source repository so if you think there are valuable updates to the documentation that others would benefit from you should feel free to put up a pull request with the updated documentation

[AhmadSafeer12345]

Helo Again,
I have a family including parents and two members and uploaded each of their VCF file in Seqr. Now when I try to load the data through dataset for the second individual in the same project it says "The following families are included in the callset but are missing some family members: 335 (15016_202926_22A). I have attached screen shots for your information. Could you please let me clarify this? Does it require a single merge VCF files of each member?

[hanars]

You need to joint call all your data in the project together. To load more data you need to joint call your previously loaded data with the new samples you want to add. You need one VCF and one elasticsearch index per project

[AhmadSafeer12345]

Hello,
I have joined call each individual and while uploading on seqr, it shows the given error.

Hail version: 0.2.122-be9d88a80695
Error summary: VCFParseError: missing value in FORMAT array. Import with argument 'array_elements_required=False'
DEBUG: 1 running tasks, waiting for next task to finish
INFO: Informed scheduler that task SeqrVCFToMTTask__seqr_reference__VARIANTS__input_vcfs_Grow_c6d1b608eb has status FAILED
DEBUG: Asking scheduler for work...
DEBUG: Done
DEBUG: There are no more tasks to run at this time
DEBUG: There are 2 pending tasks possibly being run by other workers
DEBUG: There are 2 pending tasks unique to this worker
DEBUG: There are 2 pending tasks last scheduled by this worker
INFO: Worker Worker(salt=2147527017, workers=1, host=6b0c6b2eab45, username=root, pid=25) was stopped. Shutting down Keep-Alive thread
INFO:
===== Luigi Execution Summary =====

Scheduled 3 tasks of which:

• 1 complete ones were encountered:
  • 1 VcfFile(filename=/input_vcfs/Growthmerged335/3351234.vcf.gz)
• 1 failed:
  • 1 SeqrVCFToMTTask(...)
• 1 were left pending, among these:
  • 1 had failed dependencies:
    • 1 SeqrMTToESTask(...)

This progress looks :( because there were failed tasks

===== Luigi Execution Summary =====

Could you please guide me as to how it can be solved?

[hanars]

This means you did not joint call the VCF in a way that is compatible with seqr. Please follow seqr's instructions for how to run a valid joint calling workflow and validating the expected output as defined here: https://storage.googleapis.com/seqr-reference-data/seqr-vcf-info.pdf