README.md

Python package for analysis of neuronal whole cell patch clamp electrophysiological data

• see NeuroClassify for import to Matlab from Igor Pro
• README code taken from docs/usage_example.py

First, import packages needed for these examples:

from neurospyke.classify_cell import classify_cell
from neurospyke.plot_df_utils import D1_D3_scatter_subplots
from neurospyke.query import Query
from neurospyke.utils import concat_dfs_by_index 
from neurospyke.utils import load_cell
from neurospyke.utils import load_cells 
from neurospyke.utils import reorder_df 
from tabulate import tabulate

Import electrophysiological data from Matlab.

Next, load cellular data that has been saved as .mat files.

cell_file_pattern = 'docs/ExampleCells/*.mat' 
example_cells = load_cells(cell_file_pattern)

Query the electrophysiological properties of these cells.

Each Query can include both desired criteria and properties to be calculated.

Query 1 examines an action potential spiking property, for sweeps that have .3 second current injection eliciting 5 spikes. For each response that fits this criteria (as determined by use of "response_properties"), this query assesses three different spiking properties, with the returned dataframe having the average of these calculations:

1. 'doublet_index': ratio of 2nd and 1st inter-spike intervals
2. 'delta_thresh': change in threshold over the course of the AP train
3. 'dVdt_pct_APamp__20__rising': dVdt of the AP rising phase at 20% of AP amplitude

# Query 1 
response_criteria = [('sweep_time', '<150'), ('curr_duration',.3), ('num_spikes', 5)]
response_properties = [
    'num_spikes', 
    'doublet_index', 
    'delta_thresh', 
    'dVdt_pct_APamp__20__rising'
    ]

query1 = Query.create_or_load_from_cache(
        example_cells, 
        response_criteria=response_criteria, 
        response_properties=response_properties,
        )
df1 = query1.mean_df

Query 2 examines hyperpolarization-related properties, with criteria set as having a .12 second/-400 pA current injection. This query calculates sag and rebound properties based on the average of all responses meeting these criteria, indicated by setting these properties in "calculated_cell_properties."

# Query 2
response_criteria = [('curr_duration', .12), ('curr_amplitude', -400)]
calculated_cell_properties = ['reb_delta_t', 'sag_fit_amplitude'] 
query2 = Query.create_or_load_from_cache(
        example_cells, 
        response_criteria=response_criteria, 
        cell_properties=calculated_cell_properties,
        )
df2 = query2.mean_df

Aggregate the results of both queries

combined_df = concat_dfs_by_index(df1, df2)                                                                                           
example_cells_df = reorder_df(combined_df, ['genetic_marker', 'ca_buffer', 'num_spikes'])    
print(tabulate(example_cells_df.head(), headers='keys', tablefmt='pipe')) 

	genetic_marker	ca_buffer	num_spikes	mouse_genotype	dVdt_pct_APamp__20__rising0	dVdt_pct_APamp__20__rising1	dVdt_pct_APamp__20__rising2	dVdt_pct_APamp__20__rising3	dVdt_pct_APamp__20__rising4	delta_thresh0	delta_thresh1	delta_thresh2	delta_thresh3	delta_thresh4	doublet_index	reb_delta_t	sag_fit_amplitude
020414-2	D3	Fluo5	5	D3_Ai14	243.633	171.867	218.733	242.2	158.7	0	3.22667	1.66	2.05333	2.44	1.46154	43.8	0.179389
020414-3	D3	Fluo5	5	D3_Ai14	281.717	127.683	187.233	167.733	181.417	0	5.81333	2.78333	3.02667	3.07667	2.2822	37.7	-3.3832
040915-2	D3	EGTA	5	D3_Ai14	242.7	230.967	237.8	231.467	233.4	0	2.54	1.85333	1.75667	2.24667	2.44884	33.75	-2.12921
040915-4	D3	EGTA	5	D3_Ai14	262.2	140.167	177.733	180.667	185.567	0	5.22667	3.17333	3.76	3.61333	3.17862	36.9	-2.58267
040915-7	D3	EGTA	5	D3_Ai14	241.867	134.089	179.022	196.956	190.133	0	5.40444	3.84222	4.03778	3.48444	3.00712	32.45	-2.06709

Example sweeps used for AP property calculations:

output_dir = 'docs/output/'
query1.cells[0].plot_sweeps(filepath=output_dir + '5AP_example.png')

Example response used for reb_delta_t calculation:

query2.cells[0].plot_reb_delta_t(filepath=output_dir + 'example_reb_delta_t.png')

Plot the results of these queries

comparisons = [
        ('sag_fit_amplitude', 'reb_delta_t'), 
        ('doublet_index', 'delta_thresh4'),
        ('doublet_index', 'dVdt_pct_APamp__20__rising4'),
        ('delta_thresh4', 'dVdt_pct_APamp__20__rising4'), 
        ]
D1_D3_scatter_subplots(example_cells_df, comparisons, 'EGTA',  output_path=output_dir + 'D1_vs_D3_ephys.png')

Neuronal classification based on electrophysiological properties (Clarkson et al., 2017)

cell_to_classify = load_cell('docs/example_cells/082615-8.mat')
cell_type = classify_cell(cell_to_classify, filepath=output_dir + 'example_classified_cell.png')

References:

Clarkson et al. 2017, Journal of Neuroscience: D3 Receptors Regulate Excitability in a Unique Class of Prefrontal Pyramidal Cells

Copyright 2017, Rebecca L. Clarkson. All rights reserved.