Explore samples in Gemma via the REST API

[arteymix]

During ISMB 2024, some participants who attended our poster expressed the interest of exploring Gemma's data from the perspective of samples.

Kind of use cases this would cover:

• find all control samples to be used for a null hypothesis
• retrieve expression vectors for a given set of samples that have a common characteristic

Having REST API endpoints for finding samples with certain characteristics or factors would be the bare minimum to allow these use cases.

On the backend, this means implementing the FilteringVoEnabledService interface in the BioMaterialService, adding full-text search capabilities for BioMaterials in the SearchService and exposing these functionalities in a BioMaterialWebService.

ACLs is tricky to implement because there's a one-to-many relation between a BioMaterial and a BioAssay. There might be cases of multiple BAs associated to one BM, but there shouldn't be more than one EE. For single-cell, this is likely to be more complicated because a given BM could be reused in multiple subsets.

The current workaround that we presently offer is to find datasets that have samples with specific characteristics and subsequently query the getDatasetSamples endpoint to retrieve individual samples and then retain relevant ones.