This dataset provides a comprehensive collection of multimodal data resulting from the collaboration of six leading neuroscience laboratories, collectively known as the Dendritic Consortium. The data integrates various high-resolution modalities, including in vivo calcium and voltage imaging, dendritic patch-clamp recordings, synaptic mapping, proteomics, volumetric EM-based maps, and computational neuronal models. These datasets are derived from research on Baz1a pyramidal neurons in the mouse primary visual cortex (V1). The data is publicly available to foster collaboration and accelerate neuroscience research.
The Dendritic Consortium dataset challenges the conventional model of neurons by emphasizing dendrites as active computational units, rather than passive structures. This approach is supported by a multidisciplinary combination of molecular, structural, functional, and computational data, collected using advanced neuroscience techniques. The project brings together six prestigious research laboratories: Michael Lin's lab at Stanford University, Rafael Yuste's lab at Columbia University, Jayeeta Basu's lab at New York University, Elly Nedivi's lab at MIT, and Jeff Lichtman's lab at Harvard University. Together, they aim to expand our understanding of dendritic function and its role in neuronal processing.
To fully understand dendritic function, it's essential to gather multimodal data from various experimental and computational approaches. This dataset includes calcium and voltage imaging, electrophysiology data, synaptic mapping, proteomics, volumetric electron microscopy-based maps, and computational models. These diverse techniques provide a comprehensive view of neural function across Mus musculus (focusing on Baz1a pyramidal neurons in the primary visual cortex). The dataset is available in multiple file formats, including TIFF, HDF5, GBK, ABF, PNG, PY, MAT, CSV, JSON, HOC, PKL, and ASC/SWC.
This dataset is continuously expanding, as the participating laboratories within the Dendritic Consortium regularly generate new experimental and computational data. As a result, the dataset is constantly updated.
The dataset contains multimodal files capturing:
| Data Modality | Description | Format |
|---|---|---|
| Voltage imaging files | High temporal resolution optical voltage recordings. | TIFF |
| Calcium imaging files | Optical recordings capturing calcium dynamics. | TIFF, PNG |
| Electrophysiology data | Patch-clamp recordings capturing electrical activity. | ABF, MAT, CSV |
| Synaptic mapping | Data related to synaptic connections and interactions. | CSV |
| Proteomics | Protein expression and synaptic marker data. | CSV |
| EM Volumes | 3D volumetric electron microscopy maps. | TIFF, PNG |
| Computational models | Neuronal morphologies and simulations. | SWC, HOC, PY, PKL |
| Metadata | Stored in a database accessible via a GUI. Access details will be announced. | N/A |
The dataset metadata is organized into relational tables that describe various aspects of the data. These tables offer context on data provenance, experimental conditions, parameters, and other relevant information. The metadata is accessible through the Dendritic Consortium website under the database tab. Detailed descriptions of each table's attributes (columns), including types and descriptions, are available in metadata.md. All JSON schemas are available in the json-schemas directory. Below is a summary of the metadata tables included:
| Table Name | Description | Access JSON |
|---|---|---|
| analysis | Files generated from computational/statistical analysis (graphs, reports, metrics). | analysis.json |
| behavior | Quantitative/qualitative behavioral data recorded during experiments. | behavior.json |
| cell_detection_dataset | Datasets of detected cell locations/types from segmentation tools. | cell_detection_dataset.json |
| ephys | Electrophysiology recordings, e.g., dendritic patch-clamp data. | ephys.json |
| image | Imaging files capturing calcium, voltage, or fluorescence signals. | image.json |
| micro_ct_dataset | High-resolution micro-CT volumetric anatomical images. | micro_ct_dataset.json |
| model | Computational neuronal models simulating biophysical properties. | model.json |
| mouse | Mouse specimen metadata including strain, genotype, birth date, etc. | mouse.json |
| plasmid | Plasmids used for genetic constructs (e.g., GEVIs, opsins). | plasmid.json |
| sem_dataset | Scanning electron microscopy datasets with ultrastructural details. | sem_dataset.json |
| session | Data acquisition events metadata (experimenter, date, anesthesia, etc.). | session.json |
| virus | Information on AAV vectors used for genetic delivery. | virus.json |
The dataset is publicly available on AWS S3. Data access is open and requires no sign-in, with a well-organized directory structure for easy and scalable downloads:
s3://dendritic-consortium/
├── lab-name/ ← Root folder for each participating lab
├── YYYYMMDD/ ← Date of the experimental session
├── object-type/ ← Data modality or type (e.g., image, ephys, sem_dataset)
├── version/ ← Data version (raw, processed, roi, script)
├── files/ ← Actual data files
Step-by-step examples to access, and download the multimodal dataset on AWS are available in the /tutorials directory. These tutorials demonstrate how to use both Python and the AWS CLI to interact with the data.
This dataset is a collaboration between Columbia University, Stanford University, New York University, Harvard University, and MIT.