NYU Libraries support for converting data from two NYU neuroscience labs to NWB and publishing it on DANDI
NYU Libraries funded CatalystNeuro to help two NYU neuroscience labs convert their data to NWB and publish it on the DANDI Archive. The project was coordinated by the Libraries’ Data Services department and ran in 2024 and 2025. A second goal was to train Libraries staff in standardizing and publishing neurophysiology data, so they took part throughout the process.
The two labs brought a broad mix of data. From rats performing a high-throughput Bpod task, we converted behavior, fiber photometry with optogenetics, Neuropixels recordings with Kilosort and Phy results, and DeepLabCut pose estimation, organized as four projects. From mice using sound to guide skilled movements, we converted silicon probe and Neuropixels recordings, behavior, optogenetic suppression, intrinsic signal imaging, video with SLEAP pose estimation, and audio. Each lab has an open-source conversion package and notebooks showing how to read and stream the files. The published Dandisets include 001169, the data behind the preprint “Distinct value computations support rapid sequential decisions”.
NWB conversions under this project
2 labs
David Schneider
New York University
Developed NWB conversion tools for multi-modal neuroscience data including behavioral tracking, electrophysiology recordings, optogenetic stimulation, intrinsic signal imaging, and auditory stimulation. The project includes conversion of behavioral data from .mat files, OpenEphys recordings, optogenetic stimulation parameters, intrinsic signal images, and auditory stimulation data from multiple experimental paradigms. These tools support experiments investigating how the auditory cortex integrates sound and movement information, encodes error and learning signals, and guides sensorimotor adaptation during skilled, sound-generating behaviors in mice.
Christine Constantinople
New York University
Developed comprehensive NWB conversion tools for the Constantinople lab's diverse neuroscience datasets. The conversion pipeline handles multiple data streams including fiber photometry recordings and electrophysiology data, with support for both processed behavioral data and spike sorting. The tools include specialized extractors and interfaces for different experimental paradigms, along with extensive metadata handling through YAML configuration files.