Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

parameters in PCA-ICA analysis #107

Open
NataliaFavila opened this issue Jul 25, 2022 · 6 comments
Open

parameters in PCA-ICA analysis #107

NataliaFavila opened this issue Jul 25, 2022 · 6 comments

Comments

@NataliaFavila
Copy link

Dear Biafra,

we are using Ciatah to analyse our Ca imaging data, but we are struggling finding good parameter values for the PCA-ICA cell extraction.
Can you tell me what the mu and term_tol paramters are? or point me towards any documentation or literature where they are explained?

Thank you!
Natalia

@bahanonu
Copy link
Owner

Hi Natalia,

Some explanations below. I'll add some of this information to the docs (https://bahanonu.github.io/ciatah/pipeline_detailed_signal_extraction/).

As a quick question: what type of data are you analyzing and do you have a snippet of a representative movie you are trying to analyze or can share a frame? That'll also help determine whether anything can be done on the pre-processing end to improve PCA-ICA runs or if the data will work well with PCA-ICA.

Mukamel, 2009 (mu)

The original Mukamel, 2009 (https://doi.org/10.1016/j.neuron.2009.08.009) paper describing PCA-ICA gives an explanation of mu:

image

If you look at Fig. S3, it also provides some information on the effects that varying mu 0 to 1 have on cell extraction quality (we've often found lower values, e.g. 0.1, work well):

image

Ahanonu, 2022 (mu)

We also describe it in our recent calcium imaging experiments and analysis book chapter, see section 3.15 Extraction of Neuron Shapes, Locations, and Activity Traces: https://link.springer.com/protocol/10.1007/978-1-0716-2039-7_13#Sec19.

image

term_tol

The term_tol parameter is the termination tolerance, e.g. when min difference between ICA iterations is below this value, the algorithm will exit (if it hasn't already reached max_iter).

cheers!
-Biafra

@NataliaFavila
Copy link
Author

Hi Biafra,

thank you very much for all the info, it is very helpful.
We are analysing 1p data of interneurons (obtained with GCaMP6f), so the signal-to-noise ratio is not super good.
Any recommendations will be much appreciated.

thanks!!
Natalia
frame_example

@bahanonu
Copy link
Owner

Glad to hear!

Would you be able to provide a link (e.g. Dropbox/Box) to one of the movies? I can then take a look and get back with pre-processing/cell extraction suggestions.

@NataliaFavila
Copy link
Author

Hi Biafra,

here is a link to one of the movies[movie_exmaple] (https://drive.google.com/file/d/1ROwykUCIMlNQtzyCrNj_g2ja279q2AVI/view?usp=sharing)

Our videos are already spatially filtered and spatially downsample 4x (so the movies are 270 x 270).
Thank you very much for the help!
Natalia

@bahanonu
Copy link
Owner

bahanonu commented Jul 27, 2022

Thanks, will take a look. Could you send a movie without spatial filtering? Will make it easier to test pre-processing on my end (e.g. spatial filtering for motion correction can be different than final spatial filter applied to the movie).

@NataliaFavila
Copy link
Author

Hi Biafra, unfortunately we don't have the original inscopix videos :/
Thank you so much for your help!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants