Although I perfectly agree with the goal of promoting a sample-centric approach to data management, I still believe that the origin of the data should be tracked. enpkg:FeatureList instances can be naturally grouped into "experiments". For example, RT is still useful for quality control within an experiment but not between experiments. My suggestion here would be use skos:members to describe collection of FeatureList from a single experiment. Maybe not...
Although I perfectly agree with the goal of promoting a sample-centric approach to data management, I still believe that the origin of the data should be tracked.
enpkg:FeatureListinstances can be naturally grouped into "experiments". For example, RT is still useful for quality control within an experiment but not between experiments. My suggestion here would be use skos:members to describe collection of FeatureList from a single experiment. Maybe not...I am currently investigating https://www.w3.org/TR/hcls-dataset/ as a recommendation for documenting datasets.