#811 assay harmonization - objective

[jzheng]

53 out of 175 assays specified biological objectives. It is not used in IEDB assays. However, it is useful for the functional genomics assays.

List of assays with specific biological objective (listed on the ‘objective’ worksheet):
https://docs.google.com/spreadsheets/d/1wg2Gg2OvnkX84lVeqRO0ie-tSvtEe4Pq0rtnKsveas/edit?usp=sharing

Different techniques can achieve same objective. For example, ‘transcription profiling by RT-PCR assay’ and ‘transcription profiling by array assay’ have the same objective (‘transcription profiling objective’). On the other hand, an assay could achieve multiple objectives. For example, the ‘ChIP-seq assay’ has ‘epigenetic modification identification objective’ and ‘protein and DNA interaction identification objective’. For achieving different objective, the assay process and the way to prepare samples may be differnt. It may need to create its subClass, each has a specific objective. Needs were found in annotation of Beta cell genomics data.

Specific biological objective of assay is a useful component in automatic classification of assay terms.

In addition, some cases use output approach may have maintenance issue. For example,
http://purl.obolibrary.org/obo/OBI_0001954 ChIP assay
has_specified_output some ('information content entity' and ('is about' some (
'chromatin remodeling' or 'regulation of molecular function, epigenetic' or
'sequence-specific DNA binding' or 'transcription factor binding site')))

[bjoernpeters]

This has been a long discussion, and we should decide on the need for objectives based on the overall modeling of the assay hierarchy. The proposed approach is to include an objective column in the assay modeling, which will not be mandatary to fill out. We will then assess on the global implementation of 3 alternative approaches
1) using objectives as is
2) asserting parents of assays that implictly share that objectve
3) Creating data items grouping terms that are the shared output of such assays