Issues running on cluster and locally

[HJeffery]

Issue for running locally:

Command:
./princess all -d /path/to/output/directory -r ont -e False -a minimap -s /path/to/fastq.gz -f /path/to/reference.fna -c chr15 -m -md /path/to/fast5/folder/

Error:
Invalid config definition: Config entries have to be defined as name=value pairs.

Issue for running on cluster:

Command:
./princess all -d /path/to/output/directory -r ont -a minimap -s /path/to/fastq.gz -f /path/to/reference.fna -c chr15 -m -md /path/to/fast5/folder/

Error:

Traceback (most recent call last):
  File "/path/to/working/directory/cluster/scheduler.py", line 79, in <module>
    raise Exception("Job can't be submitted\n"+output.decode("utf-8")+error.decode("utf-8"))
Exception: Job can't be submitted
Unable to run job: attribute "m_numa_nodes" is not a integer value.
Exiting.

[MeHelmy]

Thank you,

Which conda and snakemake version you are using?
What is your cluster?
Do not forget you need to decide which cluster you are using in this file princess/cluster/key_mapping.yaml
by default, it is set to system: "pbs" there is support for slurm and lsf.

The command
./princess all -d /path/to/output/directory -r ont -e False -a minimap -s /path/to/fastq.gz -f /path/to/reference.fna -c chr15 -m -md /path/to/fast5/folder/

./princess all -d /path/to/output/directory -r ont -e -a minimap -s /path/to/fastq.gz -f /path/to/reference.fna -c chr15 -m -md /path/to/fast5/folder/

Thanks,
Medhat

[HJeffery]

Hi Medhat,

I'm using a UGE cluster which is probably where the issues are coming from....

Other versions are: conda v4.10.3, snakemake v5.7.1.

Thanks for clarifying the -e is fine on its own without the False (assumed as the default was True that it required a False).

Really appreciate the help!

Heather

[MeHelmy]

Sure!