Question

Error when trying to access BiomaRt from a cluster

0

Entering edit mode

camerond • 0

@camerond-15316

Last seen 8 months ago

United Kingdom

I'm trying to run use BiomaRt on a Slurm cluster. When I run the following basic command in R (3.5.1) from my home folder it works fine.

genemart = useMart(host="www.ensembl.org",
               biomart = "ENSEMBL_MART_ENSEMBL",
               dataset="hsapiens_gene_ensembl")

However, when I send the job to the cluster it fails with the following error message:

Request to BioMart web service failed.
The BioMart web service you're accessing may be down.
Check the following URL and see if this website is available:
http://www.ensembl.org:80/biomart/martservice?type=registry&requestid=biomaRt
Error in if (!grepl(x = registry, pattern = "^\n*<MartRegistry>")) { : 
argument is of length zero
Calls: useMart -> listMarts
Execution halted

I have tried to run this from a different port (port = 443), as suggested here, but I get the same error message.

This command is the first step in quite a big list of queries that I have so I can't just run the entire script from my home folder. I can't work out why this error is occurring as I'm loading the same version of R, and the using the same version of BiomaRt (2.38.0).

Any ideas on how to get around this would be greatly appriciated.

biomaRt R cluster • 1.7k views

ADD COMMENT • link updated 4.3 years ago by Martin Morgan 25k • written 4.3 years ago by camerond • 0

0

Entering edit mode

The package developer may respond but have you also tried host = 'uswest.ensembl.org' or host = 'uswest.ensembl.org'?

ADD REPLY • link 4.3 years ago Kevin Blighe ★ 4.0k

0

Entering edit mode

Many thanks for the suggestion. I have tried it, but no joy unfortunately. Unsure why this works when I run it locally but not when sent to the cluster???

ADD REPLY • link 4.3 years ago camerond • 0

score 0 · Answer 1 · 2020-07-25

0

Entering edit mode

Martin Morgan 25k

@martin-morgan-1513

Last seen 3 months ago

United States

Probably there is something about internet access from your cluster. Try to access the list of marts directly

url = "http://www.ensembl.org:80/biomart/martservice?type=registry&requestid=biomaRt"
response = readLines(url)

response should be a character vector with contents like (abbreviated)

> response
> response
 [1] ""
 [2] "<MartRegistry>"
 [3] "  <MartURLLocation database=\"ensembl_mart_100\" default=\"1\" displayName=\"Ensembl Genes 100\" host=\"www.ensembl.org\" includeDatasets=\"\" martUser=\"\" name=\"ENSEMBL_MART_ENSEMBL\" path=\"/biomart/martservice\" port=\"80\" serverVirtualSchema=\"default\" visible=\"1\" />"
...
[10] "</MartRegistry>"

but I guess you actual receive an error response of some sort...

ADD COMMENT • link 4.3 years ago Martin Morgan 25k

0

Entering edit mode

Yes, you are right. It threw an error. I'll ask our cluster support if there is a work around. Looking into this a bit further it seems BiomaRt could not even be loaded using library(biomaRt) command so it was failing a bit earlier than I thought. Many thanks for your tip.

Error in file(con, "r") : 
cannot open the connection to 'http://www.ensembl.org:80/biomart/martservice?type=registry&requestid=biomaRt'
Calls: readLines -> file
In addition: Warning message:
In file(con, "r") :
URL 'http://www.ensembl.org:80/biomart/martservice?type=registry&requestid=biomaRt': status was 'Couldn't connect to server'
Execution halted

ADD REPLY • link 4.3 years ago camerond • 0

0

Entering edit mode

Wait, just to be sure, the package is called biomaRt, so, to load, you'd need library(biomaRt) (not library(BiomaRt))