Here we describe initial KEfED modeling efforts restricted to a set of four papers that are curated into the IEDB. This provides the initial models that we will then attempt to model in a KEfED-enabled OBI-derived formalism.

Initial papers

• PMID 9499101
  • Figures 3 and 4, Table 3 challenge assay after immunization (BCell Assay)
• PMID 19224639
  • Figure 1C and Supporting Information Table 1 challenge assay after immunization
• PMID 14671086
  • Figure 1 antigen processing assay
• PMID 16982895
  • Table 1 and Figures 1, 2, 3, 4 and 5 x-ray crystallography assay

We will initially focus on Richardson et al 1998 (PMID 9499101) to flesh out in detail.

Modeling the workflow in the paper itself

Richardson et al 1998 - Fig 3 + 4

[pdf]

In discussion with Bjoern and the IEDB team, we need to tailor he model more precisely to the existing representation in IEDB. This includes referring to these elements from IEDB.

1. Record for the paper: (http://www.iedb.org/reference/1006618)
2. Record for the assay: (http://www.iedb.org/assay/1509122)
3. Record for the epiptope: (http://www.iedb.org/epitope/53122)
4. Note that we also want to frame our representation of the assays we are using directly to the structure of the 'Assay Finder' elements in the IEDB interface.

Modeling the workflow to emulate IEDB

We redesigned this following feedback in order to match the existing design of the data in IEDB:

[pdf]

One of the goals of this modeling work is also to add a 'part-of' relationship where a subsection of a more fine-grained model can be incorporated into a single process. It is currently unclear how to effectively model that in OBI.

[pdf]

Finally, when we add representations for all the variables and measurements for this specific experiment, the design looks like this:

[pdf]