usage of OMA on a HPC cluster with slurm

Question

[Yang]

I installed the OMA.2.1.1 on a HPC cluster with slurm. AS the HPC cluster has no connection to the internet for the safety.

So when I run the software, the error is Resolving www.geneontology.org... failed: Name or service not known. wget: unable to resolve host address `www.geneontology.org' Error, (in wget) wget failed

What should I do to solve this problem?

[GenoMax]

What should I do to solve this problem?

If the cluster has no internet connection you could figure out what OMA needs from geneontology.org and see if you can provide it locally as a downloaded file.

I am tagging one of the developers adrian.altenhoff to see if he can provide any additional clues.

[Yang]

Thanks for your advice.

[Yang]

Hi @adrian.altenhoff Thanks for your kindly reply.

I copied the GOdata.drw-20180108.gz and GOdata.drw.gz to the data folder. Please see the screenshot http://ibb.co/bwgZMm

And then I use the ToyExample to do the test. But it seems do not work. See the screenshot. http://ibb.co/mSsPMm

[GenoMax]

Please use ADD REPLY/ADD COMMENT when responding to existing posts to keep threads logically organized. This comment belongs under @Adrian's answer.

[Adrian Altenhoff]

Hi @Young, if you have installed OMA standalone on the cluster, the files from the data folder need to be stored in ${HOME}/.oma/. So you should end up with something like ${HOME}/.oma/GOdata.drw.gz

[Yang]

Hi @adrian.altenhoff

Thanks for your kindly reply.

I copied the GOdata.drw-20180108.gz and GOdata.drw.gz to the data folder. Please see the screenshot. http://ibb.co/bwgZMm

And then I use the ToyExample to do the test. But it seems do not work. See the screenshot. http://ibb.co/mSsPMm

[Yang]

This time I sorted the GOdata.drw.gz and GOdata.drw-20180108.gz files in the /HOME/sdau_ssli_1/Software-W/OMA folder where I installed the OMA standalone. http://ibb.co/dhZyE6

But this doesn't help.

[Adrian Altenhoff]

no, it should be in your home/.oma folder. ls ~/.oma/ should list those files

[Yang]

I checked the /HOME/sdau_ssli_1/.oma/, and I can see the two files and data file. http://ibb.co/kbgiMm

Is it right this time? But it doesn't work, the error is wget failed.

[GenoMax]

Please use ADD REPLY/ADD COMMENT when responding to existing posts to keep threads logically organized.

ALL of these comment belongs under @Adrian's answers/comments.

[Yang]

Hi @genomax, I am sorry to disturb you. But everytime I press the 'ADD REPLY/ADD COMMENT'button, nothing has happened. Then I press the link button, the browser shows me the 'Add your answer'box. I do not know why my response belongs to your answer.

[GenoMax]

You need to have javascript turned on (white list biostars.org for scripting). Once you click ADD COMMENT/ADD REPLY a new window should open where you can type your response in.

You are not disturbing me but entering responses as new answers makes it very hard to figure out what a real answer is for someone who will find this thread by searching in future.

Answer 1

[Adrian Altenhoff]

Hi @Young,

OMA needs the ontology definition from geneontology and converts it into an internal datastructure. The simplest solution is to run the conversion step of oma (run bin/oma -c) on a machine with internet access and copy the resulting files GOdata.drw* in the newly created folder data to the cluster.

Cheers Adrian

[Adrian Altenhoff]

if you have installed OMA standalone on the cluster, the files from the data folder need to be stored in ${HOME}/.oma/. So you should end up with something like ${HOME}/.oma/GOdata.drw.gz

Answer 2

[Yang]

I coped the GOdata.drw-20180108.gz and GOdata.drw.gz to the home/.oma folder, http://ibb.co/mFOKP6,

But it doesn't work.

As I use HPC cluster with slurm, my home folder is HOME/sdau_ssli_1 and I have no root permissions. Is this the point that causes the problem?

[Adrian Altenhoff]

I don't understand where you checked. in home/.oma, HOME/sdau_ssli_1/.oma or somewhere else. It should be in a subfolder called ".oma" in your home directory. you should be able to see the files if you run ls $HOME/.oma in the slurm cluster. If this is the place where you stored them, please check with

find /HOME/sdau_ssli_1/Software-W/OMA -name "darwinit" | xargs grep datadirname

the full path where oma is checking for the GOdata.drw.gz file.