Configuration form needs a redesign

[laceysanderson]

The current configuration for this module is split out over 3 very intuitive forms. This makes it very overwhelming...

Instead, this module should have a single efficient configuration form. While we don't want to go crazy on features (e.g. ajax) this is the first impression administrators get of our module. furthermore it is critical that users configure this module properly as they can't change it once data is uploaded.

Suggested Design:

Notes:

• All of this can be done without additional themeing (for the table see: http://e9p.net/theming-drupal-7-form-elements-table)
• Currently we have a single "Database" & "Controlled Vocabulary" on the "Database & CV Terms" form. This needs to become the "Trait Vocabulary" + "Associated Database" per genus. However, this conversion is non-trivial and beyond the scope of this specific issue. Thus simply show the drop-down for the top genus for now. More work will be done as part of #10.
• Currently we have two drop-downs for setting a cvterm. The first for the cv and the second for the term. In the new design, this is a single autocomplete textfield with the format term (cv) When making the autocomplete callback, make sure it allows the match to start anywhere in the term & don't include the cv name in the search. Also, validation is needed to make sure the user selected the term from the drop-down.
• We should also provide really good defaults for the cvterms in the hopes that the user won't have to change them at all. I've started in the mock-ups although obviously the remaining lorem (ipsums) need to be replaced. I'll comment back

[laceysanderson]

Text for easy copy/paste :-)

This module requires that phenotypic traits be part of a controlled vocabulary.

• Trait Vocabulary: a container of terms where each term is a phenotypic trait that can be measured in your species of interest. This controlled vocabulary should be specific to a given genus and each term will become a trait page on your Tripal site. If you do not already have a trait vocabulary, you can create it [here] and add terms upfront and/or automatically on upload of phenotypic data.
• Associated Database: Chado requires a "database" container to be associated with all controlled vocabularies. Please select the "database" container you would like to be associated with your trait vocabulary. If needed, create one.
• Crop Ontology: our experience with breeders has led us to recommend using the trait names your breeder(s) already use in the Trait Vocabulary and then linking them to a more generic crop ontology such as those provided by cropontology.org to facilitate sharing. If you decide to go this route, you can set the species specific crop ontology here and on upload suitable terms will be suggested based on pattern matching.

Please select the appropriate vocabulary for each genus you intend to support phenotypic data for.

Once phenotypic data has been uploaded for a genus, these vocabularies cannot be changed! Please take the time to read the description above and if you have questions, submit a ticket to https://github.com/UofS-Pulse-Binfo/analyzedphenotypes/issues

Chado uses controlled vocabularies extensively to allow for flexible storing of data. As such, this module supports that flexibility to ensure that you have the ability to choose the terms that best support your data.

Please indicate the term we should use to indicate the property/relationship types specified below.

We have helpfully selected what we think are the best ontology terms below. Thus the following configuration is completely optional, although I do recommend you review our choices.

This term indicates that a given property is the associated "genus". For example, this module saves experiments producing phenotypic data as projects and the organism that experiment assays is indicated by adding a "genus" property to it.

Once phenotypic data has been uploaded for a genus, these terms cannot be changed! Please take the time to read the description above and if you have questions, submit a ticket to https://github.com/UofS-Pulse-Binfo/analyzedphenotypes/issues

[laceysanderson]

CV Terms for Defaults

• Genus: term=genus; cv=taxrank; loaded via core Tripal
• Method: term=method; cv=NCIT; URL=http://purl.obolibrary.org/obo/NCIT_C71460
• Related: term=related; cv=synonym_type; loaded via core Tripal
• Year: :-(
• Location: term=location; cv=NCIT; URL=http://purl.obolibrary.org/obo/NCIT_C25341
• Replicate: term=replicate; cv=EFO; URL=http://www.ebi.ac.uk/efo/EFO_0000683
• Data Collector: term=collector name; cv=GCP Germplasm Ontology

With the exception of those "loaded via Tripal Core", these default cvterm should be loaded when the module is enabled. Please use the following code snippet as an example, for consistency among our modules :-)

  $ontologies['NCIT'] = array(
    'name' => 'NCI Thesaurus OBO Edition',
    'definition' => 'NCIt is a reference terminology that includes broad coverage of the cancer domain, including cancer related diseases, findings and abnormalities.',
    'all terms' => FALSE,
    'terms' => array(
      array(
        'id' => 'NCIT:C86971',
        'name' => 'Modal Chromosome Number',
        'definition' => 'The most common chromosome number within a population or sample.',
      ),
      array(
        'id' => 'NCIT:C19924',
        'name' => 'Principal Investigator',
        'definition' => 'An investigator who is responsible for all aspects of the conduct of a study.',
      ),
      array(
        'id' => 'NCIT:C25714',
        'name' => 'Version',
        'definition' => 'A form or variant of a type or original; one of a sequence of copies of a program, each incorporating new modifications.',
      ),
    ),
);
foreach ($ontologies as $abbrev => $details) {
      // First add the cv.
      tripal_insert_cv($details['name'], $details['definition']);

      // Now add each term.
      foreach ($details['terms'] as $term) {
        $term['cv_name'] = $details['name'];
        tripal_insert_cvterm($term);
      }
}

NOTE: Data Collector: term=collector name; cv=GCP Germplasm Ontology is already available in our system but it cannot be assumed to always be. As such, you need to add it via your module in such a way that it is not duplicated. Make sure to test that your settings in this array match exactly how it is stored in KnowPulse.

[laceysanderson]

In Reference to year, for the moment simply don't set a default for this field. We still have some clarification from the researchers before we can move forward with this term.

To summarize where we're at:

• We're concerned about the use of year because it may not be specific enough... For example, how do we handle more then one planting in a single year (e.g. greenhouse)?
• The best definition we have come up with so far is: "The year the phenotypic measurement was taken in. If measurements for a given plot spanned the boundary between calendar years (e.g. Dec30-Jan5) then enter the year the majority of measurements were taken in."
• We are also concerned if this definition will cause confusion since the values for a given plot for some traits will have a different year then others in countries where the growing season spans the boundary between years.

Questions for the biologists:

• Define the "year" portion of "Site-Year". What exactly do you mean? Year planted in? Harvested in? data taken in?
• Is the term site year only used in field trials or does it also apply to the greenhouse?
• How do you define the site-year when the growing season spans the boundary between years?
• What about multiple plantings per year for a single experiment? How do you uniquely identify a site-year in this case?

Where do we currently use year:

• part of site-year in the bean charts
• saved during data upload
• filter criteria in datas download
• in the experiment summary pane on trait/germplasm pages
• in the main data summary page
• part of the phenotype.uniquename

Despite the concern, our current plan of action is to stick with the concept of "year". We will create a local cvterm with the definition we decide upon after consulting with the biologists (current favourite is "year measured: The year the phenotypic measurement was taken in. If measurements for a given plot spanned the boundary between calendar years (e.g. Dec30-Jan5) then enter the year the majority of measurements were taken in."). We will also add a caveat to the upload page telling the used to contact us if there is the potential for more then one "planting?" in a single year. Then we will deal with it by adding to the form and to phenotypeprop at a later date.

[laceysanderson]

Done 😀