neuropixels-analysis

To run end-to-end spike sorting and drift correction for Neuropixels data, use this Skill to automate preprocessing, motion correction, sorting, and curation into a single reproducible workflow. It processes raw Neuropixels 1.0 and 2.0 recordings directly into curated units.

Yes, you can use Kilosort4, SpykingCircus, or Mountainsort5 for Neuropixels spike sorting. The Skill integrates these sorter backends with SpikeInterface to execute post-processing and generate quality metrics for your recordings.

AI-assisted curation in spike sorting uses automated algorithms alongside rich visualizations to evaluate unit quality metrics. This approach helps you rapidly identify and filter well-separated units from noisy clusters, enabling faster and more informed curation decisions.

To export curated spike sorting results to NWB or Phy, use the built-in export options provided by the SpikeInterface framework. The Skill processes raw recordings through post-processing and quality metrics, supporting direct export to both formats for downstream visualization.

Yes, the Neuropixels analysis pipeline supports both 1.0 and 2.0 probe layouts. It leverages ProbeInterface to handle the specific contact geometry and channel mappings required for accurate spike sorting and drift correction across both probe versions.