What is the meaning of atlas_id and st_level in 1.json?

Technical
, , ,
1

Dear forum users,

A node in the hierarchy tree of the AIBS Mouse P56 1.json file looks like this:

                 {
               "id": 68,
               "atlas_id": 998,
               "ontology_id": 1,
               "acronym": "FRP1",
               "name": "Frontal pole, layer 1",
               "color_hex_triplet": "268F45",
               "graph_order": 7,
               "st_level": 11,
               "hemisphere_id": 3,
               "parent_structure_id": 184,
               "children": []
              },

Most of the displayed attributes are self-explanatory. Still, I couldn’t find the meaning of “atlas_id” and “st_level”. I have made a search starting from Class: Structure — Allen Brain Atlas API, but I didn’t find what I was looking for.

Presumably, “atlas_id” is null when no atlas dataset is available for the region of interest.
But how does this identifier relate to an atlas?

Would you be kind enough to tell me the meaning of “atlas_id” and “st_level”?
Many thanks in advance, Luc

2

Hi Luc

The ‘atlas_id’ is somewhat deprecated. It was in use early on before we had a proliferation of atlases.

The st_level is an attempt to improve atlas navigation by lining up structures with similar level of complexity.
For example root = 0
grey and fiber tracts are 1
2 contains: cerebellum, brainstem and cerebrum
and so on.

3

Dear Wayne,

Thank you very much for your prompt answer.

Then the attribute “atlas_id” is certainly something I should not pay too much attention to.

Regarding the “st_level” attribute, presumably the abbreviation of “structure level”, would you know more about the rules used to defined this number for a given region?

For instance, in the following excerpt of 1.json (AIBS P56 Mouse)

{
           "id": 315,
           "atlas_id": 746,
           "ontology_id": 1,
           "acronym": "Isocortex",
           "name": "Isocortex",
           "color_hex_triplet": "70FF71",
           "graph_order": 5,
           "st_level": 5,
           "hemisphere_id": 3,
           "parent_structure_id": 695,
           "children": [
            {
             "id": 184,
             "atlas_id": 871,
             "ontology_id": 1,
             "acronym": "FRP",
             "name": "Frontal pole, cerebral cortex",
             "color_hex_triplet": "268F45",
             "graph_order": 6,
             "st_level": 8,

we see that the “st_level” of isocortex is 5 and the “st_level” of its child region FRP is 8?

Why such a jump? What is the rational behind the assignment of “st_level” in this example? What is the rational in general?

Thanks in advance for your guidance.

Best regards,
Luc

4

Hi Luc

Sorry, I do not have any further details on how the assignments were made, and the person who could best answer this is on hiatus right now.

Best regards,
Wayne

5

Hi Wayne,

it’s been a while since the last reply, however, I am wondering if it would be possible to shed some light on Luc’s last question concerning the rules for assigning the structure levels.

Thank you in advance,
Eleftherios

6

I have been wondering about the relationship between ‘depth’ and ‘st_level’ (‘depth’ is not included in the structure graph JSON file, but it exists in some CSVs when querying the API).

It seems like most regions I would like to include in my data visualizations have a ‘depth’ of 6 and ‘st_level’ of 8. But there are a few exceptions (median eminence has a ‘depth’ of 5 and ‘st_level’ of 8).