Context
This came up in discussion with @huddlej in the context of seasonal-flu, but seems like a good feature to consider for all pathogen workflows.
When customizing workflows, users may want to include additional annotations that are not available in the default workflow. This can be done with custom metadata columns, but that would only annotate the tips. If they want annotations across all nodes, these require the node data files. We should standardize a way for users to easily customize the input node data files for the rule that runs augur export.
Possible solutions
- Use
custom_rules to import a new rule that overwrites the existing augur export rule with custom inputs.
- Use new config param, e.g.
export.node_data, that expects of list of node data files. I think this is doable with the shared resolve_config_path function that we already use for other config file paths.
- Expect users to define a
node_data_files input function (e.g. ebola/bdbv)
- Others?
Context
This came up in discussion with @huddlej in the context of seasonal-flu, but seems like a good feature to consider for all pathogen workflows.
When customizing workflows, users may want to include additional annotations that are not available in the default workflow. This can be done with custom metadata columns, but that would only annotate the tips. If they want annotations across all nodes, these require the node data files. We should standardize a way for users to easily customize the input node data files for the rule that runs
augur export.Possible solutions
custom_rulesto import a new rule that overwrites the existingaugur exportrule with custom inputs.export.node_data, that expects of list of node data files. I think this is doable with the sharedresolve_config_pathfunction that we already use for other config file paths.node_data_filesinput function (e.g. ebola/bdbv)