Status of this document

This is a work in progress, which, once finished, is supposed to be a gentle introduction into SciMesh as well as a reference for it.

It is also a “request for comments”. Thus, if you want to contribute ideas or content, or if you found a mistake, don’t hesitate to get in touch.

Todo list

Without significant ordering:

• Add further triplets to the RDF output, even it is makes it highly redundant. For example, inverse relations like “is_specified_input_of” are interesting. There are probably tons of vocabulary to investigate, especially in the OBO ecosystem.

• Create proper entities for people (operators, “currently responsible person” of a sample). Currently, only a string literal containing the login name is emitted.

• Highly JuliaBase-specific: Create proper entities for sample topics.

• The current scheme has no persistent way to identify a sample. (This may turn out to be an implementation detail as JuliaBase does have a persistent identifier for each sample.)

• Find a way to deal with sample renames.

• Currently, most predicates are home-brew (i.e., they use our URI). They should be replaced with existing, widely adopted third-party URIs wherever possible.

• We should assure that the kind of properties is clear, either by using an established predicate or by other means. Eventually, a re-using party must be able to derive that e.g.‌ a temperature is a temperature.

• Connect process classes (in other words, experimental methods) with third-party entities representing the underlying method(s).