Add support for sample replicates

[sveinugu]

From https://github.com/orgs/fairtracks/teams/fairtracks-developers/discussions/33

"The solution adopted at IHEC is that we can link two samples as being biological replicates of each other. If two experiments are technical replicates of each other, they are simply attached to the same sample."

• [ ] Implement sample replicates as in IHEC
• [ ] Update or replace the example data to make use of this
• [ ] Update TrackFind references for BLUEPRINT accordingly

[sveinugu]

See https://github.com/orgs/fairtracks/teams/fairtracks-developers/discussions/6 about the example data. The natural way to update is then perhaps to move from GEO to ENCODE ids?

[dzerbino]

I could replace with EGA Sample IDs (e.g. ERS466414)

[dzerbino]

Looking at this more in details, the current IDs are BioSample IDs, e.g.: http://identifiers.org/resolve?query=BIOSAMPLE:SAMEA2556037 I don't understand why we are discussing GEO IDs nor how we would map this to ENCODE IDs.

[sveinugu]

@dzerbino Sorry about the confusion. This is about the example JSON file in the standard, which is copied from ENCODE. Not BLUEPRINT

[sveinugu]

I should probably locate the BIOSAMPLE Ids for the examples, if I can find a link

[dzerbino]

Gotcha! See example above

[sveinugu]

I see now that this has to be postponed, as it will require that an experiment can point to more than one sample, which we do not have downstream support for currently.

@dzerbino @dtitov