ejhonglab / elena_analysis Goto Github PK

pip install -r exact-requirements.txt

1. Convert raw movie to tiff:

cd ~/flyfood
thor2tiff .

2. Draw ROI:
   • Open raw.tif in ImageJ, analyze -> tools -> ROI manager.
   • Draw one ROI around the antennal lobe for each plane.
   • Rename all ROIs to 'al'.
   • Save ROIs as RoiSet.zip in the flyfood folder.
3. Run analysis.py, this will generate the following in stimfile_dir:
   • merged_dff.svg: weighted average df/f across merged ROIs; legend: 1 flyfood/1 control -> each component; 4 flyfood/4 control -> incomplete mixture from dropping 1 component; 5 flyfood/5 control -> complete mixture.
   • x_correlation.png: correlation calculated from maximum df/f during first 3 seconds after odor onset of each pixel in the ROI; x can be combinations of smooth thresh and sorted depending on your choice.
   • single_trial_max_dffs.csv and multi_trial_max_dffs.csv if generate_csv == True: maximum df/f traces during the first 3 seconds after odor onset; single: data for individual odor components; multi: data for complete and incomplete mixture.
   • odor_list_unqiue.p: odor presentation order without repeat.
   • trial_bounding_frames.yaml: start, odor onset, and end time frames of each odor trial
4. Run heatmap.py. This generates the maximum pixel df/f for each plane (x-axis) and repeat (y-axis) during the first 3 seconds after odor onset for each odor trial (title). Images are stored in the new folder heatmap in stimfile_dir.
5. Identify glomeruli:
   • Install this.
   • Draw one ROI around each glomerulus.
   • Rename so that ROIs with the same glomerulus have the same name.
   • Save ROIs as RoiSet1.zip in the flyfood folder.
6. Run ROI.py, this will generate the following:
   • dff_movie.tif in flyfood if write_dff_movie == True: df/f movie.
   • x_ROI.png: correlation calculated from maximum df/f during first 3 seconds after odor onset of each pixel in the glomeruli ROIs; x can be combinations of smooth thresh and sorted depending on your choice.
7. Run plot_stats.ipynb in jupyter notebook. This will plot the maximum df/f traces for each odor trial for all experiments performed on the same day.

HongLab @ Caltech Dropbox/Rotation/Elena Fall 2021/2021-11-30/3

1. orn_distance.py: code for data analysis on Hallem dataset with fruit samples, HC_data_raw.csv.
2. control_mixture.ipynb: select control mixture based on pmi, correlation, and water solubility; data folder: HongLab @ Caltech/Rotation/Elena Fall 2021/data