Issue with kilosorting 4-shank probe data

[antaramaj16]

I am trying to phy sort my kilosorted data for a session where we used a 4-shank probe and I am running into an issue.

When I Kilosorted, I used this probe layout file provided and it worked because it had the same x-coordinates, y-coordinates, among other things which we used during our recording. This was the probe layout we used:

Then when I try to phy sort this kilosorted data - the spikes look fine and in the probe view I can see that different clusters are in different shanks but in the Cluster view in the column that says 'sh' there is only 0 across all of the clusters. I am wondering if there is a problem and thereby it is not registering and not assigning each cluster with a different shank id number which it should not all be zero.

Could someone help me with this please?

[nsteinme]

I think you can get the shank ID by finding the peak channel and taking its x-coordinate. Or you could sort each shank separately in kilosort. I'm not sure how to get the sh column in phy to show a meaningful number but I expect that you shouldn't need it.

…

On Thu, Oct 27, 2022 at 9:16 AM Antara Majumdar ***@***.***> wrote: I am trying to phy sort my kilosorted data for a session where we used a 4-shank probe and I am running into an issue. When I Kilosorted, I used this probe layout file provided and it worked because it had the same x-coordinates, y-coordinates, among other things which we used during our recording. This was the probe layout we used: [image: image] <https://user-images.githubusercontent.com/74184866/198343873-8119303f-4e4b-4351-b8f5-3ea0c4152e6e.png> Then when I try to phy sort this kilosorted data - the spikes look fine and in the probe view I can see that different clusters are in different shanks but in the Cluster view in the column that says 'sh' there is only 0 across all of the clusters. I am wondering if there is a problem and thereby it is not registering and not assigning each cluster with a different shank id number which it should not all be zero. [image: image] <https://user-images.githubusercontent.com/74184866/198344094-fa6e88c7-bff5-4c6a-95a2-e3dd5064bfb9.png> Could someone help me with this please? — Reply to this email directly, view it on GitHub <#517>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABZ5IP7NN3O6IF3PWACRSOLWFKTHXANCNFSM6AAAAAARQHMNCY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>