BlechCTA

Scripts used for Christina's Ph.D. thesis on neural mechanisms underlying learned and non-learned gaping.

Repository Structure

• src/ - Source code for analysis modules
  • CM_scripts/ - Core analysis pipeline scripts
    • combine_classifier_files.py - Combines Blech EMG Classifier segment files
    • create_tau_dict.py - Combines tau, spike trains, and changepoint data
    • initialize_dataframe.py - Adds important metrics to dataframe
    • extract_emg_from_transition.py - Calculates behavior frequency per session
    • extract_emg_from_transition_aggregate.py - Aggregates behavior frequency across sessions
    • neural_behavior_correlations_aggregate.py - Neural-behavior correlation analysis
  • gape_temporal_difference/ - Temporal analysis of gaping behavior
    • gapes_temporal_features.py - Feature extraction for gape analysis
  • lfp_analysis/ - Local field potential analysis
    • collect_data.py - Data collection for LFP analysis
    • cross_trial_analysis.py - Cross-trial LFP analysis
    • plot_LFP_spectrogram.py - LFP spectrogram visualization
    • data_dirs_LFP.txt - Configuration file for LFP data directories
    • notes.txt - Analysis notes and documentation
  • multibehavior_transition/ - Multi-behavior transition analysis
    • multibehavior_transition.py - Main transition analysis script
    • DEPENDENCIES.md - Dependencies documentation
    • NOTE.txt - Analysis notes and documentation
• data/ - Processed data files
  • changepoint_gapes/ - Changepoint detection data for gapes
    • gapes_cp02_df.pkl - Changepoint 0-2 dataframe
    • gapes_cp23_df.pkl - Changepoint 2-3 dataframe
• artifacts/ - Generated analysis artifacts
  • changepoint_gapes/ - Sorted features and labels (numpy arrays)
    • sorted_X.npy - Sorted feature matrix
    • sorted_y.npy - Sorted labels
  • lfp_analysis/ - LFP analysis artifacts
    • pre_stim_data/ - Pre-stimulus LFP data for individual sessions
• plots/ - Generated visualization outputs
  • changepoint_gapes/ - Gape-related plots
    • binned_temporal_evolution_features.png - Binned temporal feature evolution
    • binned_temporal_evolution_features_cutoff_200ms.png - Binned temporal evolution with 200ms cutoff
    • binned_temporal_evolution_individual_features_cutoff_200ms.png - Individual feature evolution with cutoff
    • pca_explained_variance_smoothed_binned_features.png - PCA variance explained
    • temporal_evolution_features.png - Temporal feature evolution
    • temporal_evolution_individual_features_cutoff_200ms.png - Individual feature temporal evolution
  • lfp_analysis/ - LFP analysis plots
    • median_diff_z_power_pre_post_changepoint.png - Median power difference pre/post changepoint
    • pre_post_changepoint_spectrograms.png - Spectrograms pre/post changepoint
    • taste_averaged_band_power_pre_post_changepoint.png - Taste-averaged band power
    • pre_stim_spectrograms/ - Individual session pre-stimulus spectrograms
  • multibehavior_transition/ - Behavior transition plots
    • two_test_changepoint_plots/ - Individual session changepoint visualizations

Order of Operations

Initial set-up:

1. Run each session recording through Pytau, Blech_EMG_Classifier, and BlechClust packages before running this package.
2. Run create_tau_dict.py # combines tau, spike trains, and num cps from cp model into dictionary
3. Run combine_classifier_files.py # puts all Blech EMG Classifier segments files into one dataframe
4. Run initialize_dataframe.py # Adds important metrics to dataframe

Behavior-only analyses:

1. Run extract_emg_from_transition.py # calculates frequency of each behavior across trials, plots, and saves artifacts for each session
2. Run extract_emg_from_transition_aggregate.py # plots frequency of each behavior across trials, averages across all sessions