Human Brain Project Knowledge Graph is a metadata database consisting of datasets contributed by collaborators of the Human Brain Project and curated by human curoators in order to ensure the highest standards.
The interactive atlas viewer fetches the datasets relevant to the template space and parcellation atlas selected by the user using the following conditions:
!!! note
v2.3.0 changed the way datasets are fetched from the knowledge graph.
## Species
Datasets are now fetched from the knowledge graph on a region of interest basis.
The relevant species of datasets catalogued by Knowledge Graph are obtained from the following links:
```json
{
"fieldname":"query:species",
"relative_path":[
"https://schema.hbp.eu/minds/specimen_group",
"https://schema.hbp.eu/minds/subjects",
"https://schema.hbp.eu/minds/species",
"http://schema.org/name"
]
}
```
Depending on the selected template space and/or parcellation atlas, the datasets will be filtered to include only datasets from the relevant species.
### Human
If the selected template is any of:
- Big Brain (Histology)
- MNI Colin 27
- MNI 152 ICBM 2009c Nonlinear Asymmetric
**or**, the selected parcellation is any of:
- Grey/White matter
- Cytoarchitectonic Maps
- BigBrain Cortical Layers Segmentation
- JuBrain Cytoarchitectonic Atlas
- Fibre Bundle Atlas - Short Bundle
- Fibre Bundle Atlas - Long Bundle
- Cytoarchitectonic Maps
Then datasets which have *`Homo sapiens`* as one of its species described above will proceed to the next filter.
### Rat
And selected parcellation is any of:
- Waxholm Space rat brain atlas v1
- Waxholm Space rat brain atlas v2
- Waxholm Space rat brain atlas v3
Then datasets which have *`Rattus norvegicus`* as one of its species described above will proceed to the next filter.
### Mouse
And selected parcellation is any of:
- Allen Mouse Common Coordinate Framework v3 2017
- Allen Mouse Common Coordinate Framework v3 2015
Then datasets which have *`Mus musculus`* as one of its species described above will proceed to the next filter.
## Selected template space and parcellation atlas
The datasets are then filtered based on the selected template space and parcellation atlas.
The dataset may satisfy either conditionals to be presented to the user.
### Template space
The reference space associated with datasets are queried with the following querying links:
The dataset is considered relevant if the stripped `fullId` attribute[^1] of any of the reference spaces matches to:
[^1]:`fullId` is a URI, which in the case of Human Brain Project Knowledge Graph, always starts with `https://nexus.humanbrainproject.org/v0/data/`. Stripping the domain allows for easier comparison.
| Selected template space | fullId |
| --- | --- |
| Big Brain (Histology) | minds/core/dataset/v1.0.0/a1655b99-82f1-420f-a3c2-fe80fd4c8588 |
A dataset is considered relevant if **both** of the following conditionals are true:
#### Parcellation name
If the name of the selected parcellation in interactive atlas viewer matches exactly with either name of any of the `parcellationAtlas`, or any of its aliases listed below