Clustering kernels - Githubissues

AllenInstitute / visual_behavior_glm

GLM for ophys df/f during visual behavior

Other

2 stars 0 forks source link

Clustering kernels #201

Closed alexpiet closed 2 years ago

alexpiet commented 3 years ago

Iryna has been clustering cells based on their dropout scores. In addition we should cluster cells based on their functional shapes by clustering the kernels themselves.

As a first step, I should do PCA on the matrix of all kernels for a given feature.

[x] What is the dimensionality of the responses?
[x] Visualize the trajectory in space for different cre lines, or different subsets of the same cre line

matchings commented 3 years ago

clustering kernels is a good idea, would be interesting to see if there are groups of cells with different dynamics for different features

alexpiet commented 2 years ago

all-images_by_experience hits_by_experience hits licks_by_experience licks misses_by_experience misses omissions_by_experience pupil_by_experience pupil running_by_experience running