Skip to content

Latest commit

 

History

History
46 lines (31 loc) · 2.35 KB

README.md

File metadata and controls

46 lines (31 loc) · 2.35 KB

Calcium imaging

Experimental procedure

Raw data transfer and conversion to tiff files

  • use ssh to transfer the raw data

  • use the image block ripping utility from Bruker to convert the raw data to tiff files. Download the program from your Prairie version, see https://pvupdate.blogspot.com/ (e.g. Prairie 5.5 at the following link).

Registration and Cell detection

The pipeline relies on Suite2P. Read the documentation at http://mouseland.github.io/suite2p.

A minimal interface allow to launch the Suite2P in the background with presets:

Those settings are set by modifying the default options (see ops0 in process_xml.py file) in the preprocessing.py file, we modify the keys with a dictionary of the form:

PREPROCESSING_SETTINGS = {
    'GCamp6f_1plane':{'cell_diameter':20, # in um
                      'sparse_mode':False,
                      'connected':True,
                      'threshold_scaling':0.8},
    'NDNF+_1plane':{'cell_diameter':20, # in um
                    'sparse_mode':True,
                    'connected':True,
                    'threshold_scaling':0.8},
}

Each entry will be a default settings appearing in the GUI.

N.B. we extract the available information form the xml Bruker file, see here an example file.

Preprocessing and analysis

The fluorescence variations "dF/F0" were computed as followed. "dF" was the raw fluorescence corrected by the neuropil. "F0" was the sliding raw fluorescence (not corected by the neuropil), this setting impedes that the few ROIs with weak intrinsic signals and high neuropil signal would ge a F0 close to 0 and thus potentially high "dF/F0" values. This settings leads to relatively low values for the "dF/F0" variations (compared to the setting where "F0" also has the neuropil substraction).

The preprocessing step and some analysis are illustrated in the demo notebooks.