Replace "NeuronTracing" with "Reconstruction" activity in prov template

[annakristinkaufmann]

Reconstruction from imageStack and reconstruction from slice should be aligned by using the same activity naming.

[MFSY]

Before adressing this issue, I want clear answers and comments on the following:

1- Schemas are usually defined for types defined in a given vocabulary (through targetClass). If a NeuroTracing activity is a specific Reconstruction activity then they should be linked by a subClass relation in the bbp core vocabulary independently of the schemas which are used to constrain them. But if a Neurotracing activity is yet an other synonym for Reconstruction (i.e all Reconstruction activities can be safely considered as NeuroTracing activities) then: a) we should keep a unique name (Reconstruction) and a single schema for both in all of our prov templates b) assert in the bbp core vocabulary that NeuroTracing is a synonym of Reconstruction activity.

2- In case NeuroTracing activity is a subclass of Reconstruction activity, we should have two different schemas for both types only if we think the set of metadata to describe them are different. If they are described by the exact set of metadata, we should have one schema for both.

[srjimenez]

Neuron tracing, reconstruction

After a discussion with Anna and after looking into the some sources we concluded that "neuron tracing" is used as synonym of "neuron reconstruction", these terms are sometimes both used in a single page (https://goo.gl/oLa54b) I think we should also take into account what scientist and authorities in the subject are already using:

• Allen -> 3D neuron reconstruction (http://celltypes.brain-map.org/mouse/experiment/morphology/474626527)
• Neuromorpho -> reconstruction (http://neuromorpho.org/)
• LNMC -> reconstruction On the other hand, "cell reconstruction" does not refer only to morphology reconstruction, it also refers to 3D cell reconstruction from images (https://goo.gl/zKAkJ1).

In any case, whatever term we decide to use we should use it consistently (i.e. update also our ontologies).

[MFSY]

Thank you for this comment. If I understand correctly: "neuron tracing" is used as "neuron reconstruction" from various sources.

So I would suggest to:

• pick one name to represent this type of activity ("Reconstruction", "NeuronReconstruction", "Reconstruction", ...).
• use a single schema for both
• assert in the bbp core vocabulary that "neuron tracing" is a synonym of "neuron reconstruction" activity.

[srjimenez]

We cannot really use a single schema due to the "used" entities that are not the same. -For the reconstruction from slices we have a "used" "LabeledCell". -For the reconstruction from image stack we will have a "used" "ImageStack".

[MFSY]

I see. What else differentiate a reconstruction for image stack and reconstruction from slices and that are not part of the protocols description (which I guess will be different) ? If the only difference is the type of used entities then we can add a "sh:or" or "sh:xone" clause in the Reconstruction activity schema.

Thoughts ?

[srjimenez]

The activity named "neuron tracing" or "cell reconstruction" can be done using a brain slice or an image stack. This activity always uses the neurolucida software in a manual or semi-automatic mode. I would suggest that we align the way we name these activities since naming it "neuron tracing" when it has as an input an image or "cell reconstruction" when it has as an input a slice doesn't make a lot of sense.