Couple of issues with repositories.

[saeid-p]

• In its current version, dataMED appears to have over 20 different data sources representing 65k data sets and 10 different “data types”. This number is only a small fraction of the number of publicly available databases as reported in the annual Nucleic Acids Research database issue (see next paragraph).

• For example, although the link to data types appears to divide the results into some functional categories, it is not clear that the choice of these categories is based on data definition standards or ontologies. Therefore, a standard vocabulary and ontology-driven data model that fully captures the scientific use cases should precede the development of a Data Discovery Index (DDI). Moved to #231

[saeid-p]

• We agree that the repository number is limited currently. While we are working to increase DataMeds’ content, please do bear in mind that DataMed is currently a prototype and will not in its present form encompass all data resources yet.

• We also have felt the requirement for a data type ontology. The bioCADDIE team has initiated work in this area. We would like to clarify that the metadata model built by the bioCADDIE team is based on various use cases (provided by multiple stakeholders) and encompassing existing models in use by the community. Please see https://biocaddie.org/group/working-group/working-group-3-descriptive-metadata-datasets for a detailed description. The model is currently being refined based on feedback from the core technology development team as it gets implemented for multiple repositories.

[saeid-p]

• If the meaningful use of data is the objective, why is the focus on a list of data sources rather than on biological concepts? For example, should every user be familiar with GEO or Array Express or GEMM or LNCS?#215
• Why are only selected data sources provided? #216
• There does not appear to be a categorization of these data sources that is intuitively useful. A new issue has been created for each of the above questions.

[saeid-p]

• Not every user is expected to be familiar with GEO, Array Express, etc. This is what DataMed is meant to assist users with - letting them know what repositories (even if they are unknown to the user) have potential data of interest.
• The repositories are prioritized for ingestion by DataMed on the basis of criteria defined by a working group organized by bioCADDIE (for details, please check https://biocaddie.org/group/working-group/working-group-6-criteria-repository-inclusion-standards-interoperability). The group utilizes a priority score (developed and implemented by a working group in bioCADDIE: https://biocaddie.org/group/working-group/working-group-11-outreach-meeting-repository-operators) to decide the data sources that will be ingested.
• The data sources are currently classified based on the type(s) of data that are available from each of them. We welcome any suggestion on better ways to organize the data resources.

[saeid-p]

What is DataMed’s definition of a data type?

• bioCADDIE team members are currently working to develop a comprehensive data type ontology which will provide definitions of types of data that should cover the breadth required for DataMed. Moved to #217

[saeid-p]

Although data types are shown, the underlying data model is not clear; for example, what is

the relationship between SRA, Phenotype, and Gene Expression?

• The data model can be found in the DATS documentation (https://biocaddie.org/group/working-group/working-group-3-descriptive-metadata-datasets) Moved to #218

[saeid-p]

Why are these considered different data types? Why are mouse phenome database and dbGAP under the same data type?

• Data type ontology is currently being worked as a refinement of DATS 2.1 and The various relationships will be clear once the is generated. Moved to #219

[NinaSalimi]

The issue has been addressed by bioCaddie team.