Kilosort v4.0.6

Along with miscellaneous bug fixes, this release allows some sections of the GUI to be resized. After installing, you will probably notice that the proportions have changed. You can customize this yourself by clicking and dragging on the handles shown in the image below (circled red). Any resizing will be saved and restored the next time you open the GUI.

Full Changelog: v4.0.4...v4.0.6

Kilosort 2 patch1

This release fixes a bug for the spike holes problem (issue #594). The default value of the NT parameter has changed to 65,472 in configFile384. You need to use these defaults to fix the spike holes problem (or the rules specified in configFile384 for NT and ntbuff). Please recompile the CUDA functions with mexGPUall.

The release will be updated if further modifications are made (patch2, 3 etc).

Note: this aligns the changes in Kilosort 2.0.2 to those in 2.5.2 and 3.0.2.

Kilosort v4.0.4

Sorting should now work for multi-shank probes without needing to vertically stack the shanks.

This release adds distance gating for universal template positions, to match the behavior of previous versions of Kilosort. By default, templates will be excluded if they are farther than max(dmin, dminx) microns from the nearest active site. This can be adjusted manually by changing the new max_channel_distance parameter in "Extra settings." Universal template placement can be previewed in the GUI by checking "Universal Templates" under the probe plot. Note that changing any of dmin, dminx or max_channel_distance will change the positions of the universal templates, but this will not show on the plot until you change one of the other options or click "Preview Probe" again. The probe plot also includes axes to show scale, and a "True Aspect Ratio" checkbox to switch to a physically proportional representation.

Additionally, center points for template groupings are now determined in a way that should work well for multi-shank probes, removing the need for artificially stacking shanks on top of one another. The placement of these grouping centers can also be previewed in the GUI by checking "Grouping Centers" under the probe plot. You should see one column of centers on each shank. If this is not happening, or if you're using a 2D array, you may need to specify the new x_centers parameter in "Extra settings."

Full Changelog: v4.0.3...v4.0.4

Kilosort 3 patch1

This release fixes both bugs for the spike holes problem (issue #594). The default value of the NT parameter has changed to 65,472 in configFile384. You need to use these defaults to fix the spike holes problem (or the rules specified in configFile384 for NT and ntbuff). Please recompile the CUDA functions with mexGPUall.

The release will be updated if further modifications are made (patch2, 3 etc).

Note: the previous release 3.0.1 had incorrect spike times and only fixed one of the two bugs.

Kilosort 2.5 patch1

This release fixes both bugs for the spike holes problem (issue #594). The default value of the NT parameter has changed to 65,472 in configFile384. You need to use these defaults to fix the spike holes problem (or the rules specified in configFile384 for NT and ntbuff). Please recompile the CUDA functions with mexGPUall.

The release will be updated if further modifications are made (patch2, 3 etc).

Note: the previous release 2.5.1 had incorrect spike times and only fixed one of the two bugs.

Kilosort v4.0.3

This release adds pc_features.npy and pc_feature_ind.npy for use with feature view in Phy. Additionally, the GUI backend was updated from PyQt5 to QtPy, which is compatible with PyQt5, PyQt6, and PySide. This should result in fewer installation issues. If you previously installed Kilosort from source, you will need to reinstall after updating the code to update your GUI dependencies.

Several bug fixes are included as well:

• dminx will no longer be determined automatically, the default value is 32. This removes a bug with that behavior related to checkerboard electrode layouts. You may need to update the value in the GUI under "extra settings," or click on "reset GUI," to update this.
• Setting n_pcs to values other than 6 will no longer cause a crash.
• Fixed bug that could cause negative spike times to be detected for first batch.
• Added informative errors related to dminx and electrode spacing.
• Misc fixes to GUI behavior.

Full Changelog: v4.0.2...v4.0.3

Kilosort 4

Initial release of Kilosort4, linked with paper.

Kilosort 3.0

UPDATE 3/8/24: this version has the "spike holes" bug from issue #594 and is only left here as a reference. Please use the new release above "Kilosort 3.0 patch0" where the main bug has been fixed. We will continue to investigate if other bugs are left that produce this behavior.

This is the last version of Kilosort 3.0 and the last Kilosort version in Matlab. No further support or code changes for Kilosort 3 will be provided beyond this release.

Kilosort 2.5

UPDATE 3/8/24: this version has the "spike holes" bug from issue #594 and is only left here as a reference. Please use the new release above "Kilosort 2.5 patch0" where the main bug has been fixed. We will continue to investigate if other bugs are left that produce this behavior.

Kilosort 2.0

This release is the final Kilosort 2.0 build, before moving on to v2.5. A new drift correction algorithm is introduced in v2.5, which makes obsolete the drift tracking approach from v2.0. In addition, the two algorithms are incompatible and we are dropping support for drift tracking from v2.5 onwards. Users of Neuropixels 1.0 and 2.0 should transition directly to Kilosort 2.5.

Nonetheless, we expect some users will get better performance from v2.0, at least initially. Recording configurations that may benefit more from v2.0 than v2.5 are: tetrodes, single-channel recordings such as from a Utah array, probes with >30um vertical spacing.

To learn more about Kilosort 2.0, read the supplementary material and figures in Stringer, Pachitariu et al, Science 2019, and the wiki page of the v2.0 release. The following talks may also be useful: talk1, talk2.