Using ISA-API render more flexible /easier to create custom "configuration file" ?

[yvanlebras]

Dear ISAteam,

Hope all is ok on your side.

Just want to know if using ISA-JSON or ISA-API implementations help using custom configuration files to test using ISA on ecology fields creating new types of protocols and technologies notably ?

Best, Yvan

[proccaserra]

Hi @yvanlebras nice to hear from you !

The interplay between custom ISA-configurations and the ISA-API somehow different from what could be done with ISAcreator (the java desktop tool).

To ensure smooth integration of new ISA-configurations with the ISA-API component, the ideal course of action would be to discuss with us what your needs are so we can ensure the code can deal with extensions.

For instance we are aware that some groups have significantly changed ISA configurations, adding new fields/headers unknown to the ISA-API and were surprised that the ISA parser complained about it.

Having said that, if the ISA configurations are compliant with the ISA specifications, you should be able to add new assays specifying custom "measurement types" and "technology types".

• [ ] If the technology types are known to ISA (e.g mass spectrometry, nucleic acid sequencing), then those new assay definitions should used the associated headers.
• [ ] if the technology types are new/unknown, then the 'generic' workflow (using Assay Name, Raw Data File) , should be used.
• [ ] New protocol types: adding new ones for new domains is fine, but using different names for specific protocol types, which are needed to ensure conversion to other formats (e.g. NCBI SRA), will loose the ability to convert using the ISA-API.

Happy to have a call to discuss the ecology use cases . Those can be challenging I reckon, remembering discussions around folding information into ISA "Source" or "Sample".

Hope all is well.

BW.