First Question,
Where in the Galaxy file system are the parallelism modules defined? I.e. where can if learn how galaxy handles parallelism?
Second - Real - Question,
Given the following tag:
<parallelism method="multi" split_inputs="interval_file" split_mode="to_size" split_size="1" shared_inputs="normal,tumour,config" merge_outputs="snvs,indels"></parallelism>
How does galaxy handle multiple merge_outputs. I keep getting an error saying that galaxy is passing three inputs to my merge function I defined for VCF Files (which only takes two).
Shouldn't Galaxy be calling this function twice with the respected inputs accumulated for each subprocess? It seems like it is passing the following:
merge(split_inputs, ... ) where the ... is the values found in the tag set. THIS DOES NOT SEEM GOOD.
Should have the following:
merge(snvs_split_inputs, snvs_output) & merge(indels_split_inputs, indels_output)
Please help me with this, I require multiple outputs, and the parallelism works, except for this merge step.