Amd feature update

[arpaddanos]

This pull request has updates to add Features to the Curating and Data Model sections of the docs

[kkrysiak]

This title will be displayed underneath the figure. Both this and the next figure have the same title and obth labeled as "Figure 1". It looks like you have built the caption (Figure legend) embedded in some of these photos.

[kkrysiak]

See source page: https://civic.readthedocs.io/en/latest/_sources/curating/genes.rst.txt for this page https://civic.readthedocs.io/en/latest/curating/genes.html The figure caption comes from the command below (not embedded in the image)

:title: The CIViC gene page shows a curator Description and a list of Molecular Profiles containing variants assoicated to the gene

[obigriffith]

It looks like this problem persists throughout the new docs. For example Figures 3 and 4 also have the same Figure 1 title. Figures 5 and 6 share the same Figure 2 title, etc. The titles don't seem to match the figure contents.

[obigriffith]

Very nice work on these docs. The Figure titles and alt text all need to be updated. Otherwise I made just minor editorial suggestions.

[obigriffith]

Suggested change

Features are added to CIViC primarily by the Add Evidence functionality. This functionality allows the curator to see if the Feature they want to curate for is already in CIViC, and additionally if the Variant associated to the Feature is already in the CIViC knowledgebase. When the Add Evidence functionality is used, the curator sees the Molecular Profile fields that can be curated. Recall that a Molecular Profile (MP) consists of one or more Feature-Variant pairs. Simple MPs hold one Feature-Variant pair (e.g. BRAF V600E), and Complex MPs are combinations of two or more Feature-Variant pairs (e.g. BRAF V600E AND MSI High). When adding a new Feature, the curator first chooses the Feature type (Gene, Factor, or Fusion). The curator then types in the adjacent field, and typeahead functionality searches through the existing Features to see if the desired Feature already exists. Typeahead results are displayed below the field. If the typeahead does not yield the desired Feature, the curator is prompted to make a new Feature, as shown below. Once the new or existing Feature is chosen, a similar process is followed for adding a new Variant associated to the Feature, to create the Feature-Variant pair. For Complex MPs, this process is repeated for each element of the complex MP (See section on curating MPs).

Features are added to CIViC primarily with the Add Evidence functionality. This functionality allows the curator to see if the Feature they want to curate for is already in CIViC, and additionally if the Variant associated to the Feature is already in the CIViC knowledgebase. When the Add Evidence functionality is used, the curator sees the Molecular Profile fields that can be curated. Recall that a Molecular Profile (MP) consists of one or more Feature-Variant pairs. Simple MPs hold one Feature-Variant pair (e.g. BRAF V600E), and Complex MPs are combinations of two or more Feature-Variant pairs (e.g. BRAF V600E AND MSI High). When adding a new Feature, the curator first chooses the Feature type (Gene, Factor, or Fusion). The curator then types in the adjacent field, and typeahead functionality searches through the existing Features to see if the desired Feature already exists. Typeahead results are displayed below the field. If the typeahead does not yield the desired Feature, the curator is prompted to make a new Feature, as shown below. Once the new or existing Feature is chosen, a similar process is followed to select a new or existing Variant, associated to the Feature, to create the Feature-Variant pair. For Complex MPs, this process is repeated for each element of the complex MP (See section on curating MPs).

[obigriffith]

It looks like this problem persists throughout the new docs. For example Figures 3 and 4 also have the same Figure 1 title. Figures 5 and 6 share the same Figure 2 title, etc. The titles don't seem to match the figure contents.

[obigriffith]

Suggested change

Like other CIViC entities, curators can click the Revise button in the upper right of the Feature’s page to create a revision, as seen in the images below. This will open up a set of fields which can be curated for the Feature. The fields that are available will be dependent on the type of Feature chosen (Gene, Factor, or Fusion). Note also that MPs consist of Feature-Variant pairs, and Variants also have curateable fields which depend on the Variant’s Feature type. Revising Variants is very similar to Revising Features, as each Variant has its own page, with a Revise button accessible on the upper right of the screen.

As with other CIViC entities, curators can click the Revise button in the upper right of the Feature’s page to create a revision, as seen in the images below. This will open up a form with a set of fields that can be curated for the Feature. The fields that are available depend on the type of Feature chosen (Gene, Factor, or Fusion). Note also that MPs consist of Feature-Variant pairs, and Variants also have curateable fields which depend on the Variant’s Feature type. Revising Variants is very similar to Revising Features, as each Variant has its own page, with a Revise button accessible on the upper right of the screen.

[obigriffith]

See comments above about alt/title text not matching the figure content

[obigriffith]

See comments above about alt/title text not matching the figure content

[obigriffith]

Suggested change

Features are used to classify different groups of biomarkers with clinical relevance for cancer to support curation of structured data relevant to that type of biomarker. Features in CIViC are currently Genes, Fusions, and Factors, with other Features under development and planned for the future. For example, Genes are specific regions of the genome, named using HUGO nomenclature, and associated with structured data such as refseq or ensembl transcripts, aliases, and protein structures. The original CIViC Variant data model was built entirely around Genes, and the majority of CIViC documentation is based on curation of variants directly associated to the Gene Feature class. But with that in mind, the data model for other Features broadly follows many aspects of the data model for Genes, including the relationship between Genes and Variants as seen below.

Features are used to classify different groups of biomarkers with clinical relevance for cancer to support curation of structured data relevant to that type of biomarker. Features in CIViC are currently Genes, Fusions, and Factors, with other Features under development and planned for the future. Genes are specific regions (loci) in the genome, named using HUGO nomenclature (from Entrez Gene), and associated with structured data such as Ensembl transcripts, aliases, and protein structures. The original CIViC Variant data model was built entirely around Genes, and the majority of CIViC documentation is based on curation of variants directly associated to the Gene Feature class. But with that in mind, the data model for other Features broadly follows many aspects of the data model for Genes, including the relationship between Genes and Variants as seen below.

[obigriffith]

See comments above about alt/title text not matching the figure content

[obigriffith]

See comments above about alt/title text not matching the figure content

[obigriffith]

Suggested change

Features generalize a fundamental relation between a Gene and its Variants, so that for any instance of a Feature, multiple different variants may be associated with it. For example, a Fusion (Fusions are a class of Feature in CIViC) such as EML4::ALK can be associated with multiple Variants. In the case of Fusions, Variants may include different exon combinations, such as e6::e20 (EML4 exon 6 fused to ALK exon 20) and e20::e20, or the representative Fusion bucket variant when the specific breakpoints are not known (see section on Fusions and sections on other Feature types for more in-depth examples). With this generalization of the Gene-Variant data model to Features, it is especially useful to think of Variants in CIViC in terms of Feature-Variant pairs (e.g. BRAF V600E, HPV Positive, EGFR Mutation, EML4::ALK e6::e20, v::ALK Fusion). An overview of different Features in CIViC which are either implemented or under development is shown in the figure below:

Features generalize a fundamental relation between a Gene and its Variants, so that for any instance of a Feature, multiple different variants may be associated with it. For example, a Fusion (Fusions are a class of Feature in CIViC) such as EML4::ALK can be associated with multiple Variants. In the case of Fusions, Variants may include different exon combinations, such as e6::e20 (EML4 exon 6 fused to ALK exon 20) and e20::e20 (EML4 exon 20 fused to ALK exon 20). Alternatively, the representative "Fusion" bucket/categorical variant is used when the specific breakpoints are not known (see section on Fusions and sections on other Feature types for more in-depth examples). With this generalization of the Gene-Variant data model to Features, it is especially useful to think of Variants in CIViC in terms of Feature-Variant pairs (e.g. BRAF V600E, HPV Positive, EGFR Mutation, EML4::ALK e6::e20, v::ALK Fusion, etc). An overview of different Features in CIViC which are either implemented or under development is shown in the figure below:

[obigriffith]

See comments above about alt/title text not matching the figure content