Skip to content

Release Notes (0.14.3)

dyf edited this page Oct 19, 2017 · 4 revisions

Software Release Notes

The 0.14.3 release coincides with the first release of human data and models in the Allen Cell Types Database and a complete requantification of structure unionize records in the Allen Mouse Brain Connectivity Atlas based on a new revision of the Common Coordinate Framework structure ontology and voxel annotations.

Users of the CellTypesCache can filter for cells based on the species of the cell’s donor using the species argument of get_cells(). Examples of this are shown in the metadata filtering section of the example Jupyter notebook.

The Allen Mouse Brain Connectivity Atlas contains over 350 new data sets and structure unionize records have been completely reprocessed with updated 3D annotations of the Common Coordinate Framework. The structure ontology contains new structures, with subcortical annotations having changed the most. The MouseConnectivityCache get_annotation_volume method will by default return a new volume by default. You can choose which version of annotations you would like using the ccf_version argument to the MouseConnectivityCache constructor.

To access new experiments and unionize records, you will need to remove a number of files in your manifest directory so that MouseConnectivityCache will know to download the new copies:

  • MouseConnectivityCache manifest JSON
  • experiments.json
  • structures.json
  • structure_unionizes.csv (one per experiment within experiment subdirectories)

You can then call the MouseConnectivityCache get_experiments, get_structure_tree, and get_structure_unionizes methods to download the files above.

Data Release Notes

The latest updates to the online public resources available through the Allen Brain Atlas data portal were released on October 19, 2017. The Allen Cell Types Database now contains cells from human cortex, in addition to more cells from the mouse visual system and other cortical areas. A new version of the Allen Mouse Common Coordinate Framework has been released, with more detailed anatomical structure delineations in the Allen Mouse Brain Reference Atlas, and new data and tools in the Allen Mouse Brain Connectivity Atlas.

Highlights include:

New data and models: Allen Cell Types Database

  • Cortical cells from human medio-temporal gyrus (MTG) with electrophysiology recording data (279) and morphological reconstructions (107)
  • Immunohistochemical images from human MTG
  • RNA-sequencing data from 15,928 human MTG cells
  • Human cell models: GLIF (157 cell models) and perisomatic biophysical (43 cell models)
  • Mouse cells with electrophysiology and GLIF models (80), including data for two additional Cre lines (Chrna2-Cre_OE25 and Nkx2-1-CreERT2)

Updated reference atlas resources

  • Integrated portal for accessing all reference atlases in the Allen Brain Atlas portal
  • Updated voxel annotations in the Allen Mouse Common Coordinate Framework
    • Anatomical 3D annotations of 672 brain structures, including:
      • 241 structures in isocortex
      • 341 non-cortical gray matter structures
      • 90 fiber tract delineations
    • Annotated coronal slice views in the reference atlas viewer

New data and tools: Allen Mouse Brain Connectivity Atlas

  • Anatomical structure quantifications on brain connectivity have been recomputed using the new CCF annotations
  • Addition of 374 new anterograde projection data sets, for a total of 2,911 including:
    • 147 data sets in Cre driver mice (brain-wide injections)
    • 131 data sets in Cre driver mice (visual-area injections with ISI maps)
    • 96 target-defined data sets (brain-wide and visual-area injections with ISI maps)
  • Side-by-side comparison of 2-photon images and annotated CCF slices in the image viewer.

Note to users of the Allen Mouse Brain Connectivity Atlas: While reprocessing data using the new CCF annotations, we encountered errors on 5 experiments (IDs 112672974, 113036264, 157769139, 159552290, and 272829024). We have temporarily removed these experiments from search results until we resolve the issue. We apologize for the inconvenience and will alert users on this list when the data is available again.

Allen SDK users should be aware that as a result of reprocessing connectivity data for the new CCF annotations, all cached structure summary CSV files and annotation volumes should be re-downloaded. To do so, rename the directory containing your Connectivity manifest and re-run your existing scripts.

Note to users of the Allen Brain Observatory – Visual Coding: We continue to improve the quality control standards used for data generation and analysis. In the process of reviewing existing data using more rigorous standards, we identified 18 experiment containers that no longer meet our quality criteria. All files and calculated cell metrics for these experiment containers are still accessible from the Allen SDK, however the data will no longer appear via searches on the web portal. To download the updated failure indicators for these experiments, remove local copies of cell_specimens.json, ophys_experiments.json, and cell_specimens.json files and re-run existing scripts. Specific reasons for why individual experiment containers failed quality control review are summarized below:

Experiment Container ID Reason for failure
511506664 Z drift for sessions B and C, targeting for session A
511507144 Targeting for session A, hardware failure for session B (encoder data)
511507602 Z drift for sessions A and B
511510729 Tissue integrity
511510733 Z drift for sessions A and C
511510758 Z drift for session C
528799602 Photobleaching for session C
529770662 Z drift for sessions A and C
531100608 Z drift and targeting for session A
531823088 Z drift and photo bleaching for session A
532233174 Z drift for sessions B and C
547476353 Z drift for session C
550922597 Targeting for session A, Z drift for session B
551412362 Z drift for session C
553007101 Z drift for session C
553452260 Z drift for session C
560363325 Targeting for session A
560820973 Animal stress and hardware issues for sessions B and C