Python library for the analysis of amperometric single-channel nanopore recordings.

Introduction

A python package for analyzing electrophysiology and other things by Florian Lucas. (Also Matthijs Tadema)

Main features

• Load data from single-channel nanopore electrophysiology (electrophys)
• Automatic detection of baseline currents
• Threshold search detection of events
• Determination of Excluded currents (I_ex), its standard deviation (I_exSD) and dwell time

Usage example(s)

Importing module classes:

from holeypy import Trace, filters
from holeypy.analysis import Levels, Events, Features

Loading signal trace(s) and changing the active trace

# Access sample data
import pkg_resources
from pathlib import Path
sample_fname = Path(
        pkg_resources.resource_filename(
            __name__,
            "data/ProteinDigest.abf")
    )

# Load trace(s) from sample data
signal_data = Trace.from_abf(sample_fname)

# Set active trace
signal_data.set_active(1)

Setting a signal cut-off, the sample data contains a protocol of switching voltage which we want to trim

# Set the starting (t0) and end (t1) cut-off in seconds
t0 = 1.8
t1 = 5.8
signal_data.set_trim(t0=t0, t1=t1)

Adding filters to signal data

filter_frequency = 1000
signal_data.add_filter(filters.gaussian, Fs=filter_frequency)

Finding levels, which can be done automatically using Levels, or manually using set_levels. Note: Levels returns (0, 0) if no levels were found.

# Automatically find levels
levels = Levels(signal_data).run()

# Manually setting levels
baseline_current = -116
baseline_error = 4
signal_data.set_levels(baseline_current, baseline_error)

Minimal dwell time cut-offs are important as it is an easy way to discard noise. It is also optional to allow for outliers (default=2 data points)

# Set the minimal dwell time (in seconds) for each event
signal_data.set_dwell_time_cutoff(4e-4)

# (optional) set the number of outlier data points before ending an event (default=2)
signal_data.set_event_skip(2)

Finding events and features using threshold search

# If levels were not set manually, this will use Levels(signal_data) to determine them.
event_analysis = Events(signal_data)
event_analysis.run()

# The events are returned as a list of named tuples with fields: 
#   baseline_start: (int) datapoint where baseline started
#   baseline_end: (int) datapoint where baseline ended
#   event_start: (int) datapoint where event started 
#   event_end: (int) datapoint where baseline started 
#   t0: (int) offset
events = event_analysis.events

# The features are returned as a list of named tuples with fields: 
#   baseline_current: (float) baseline current
#   baseline_sd: (float) square root of the variance of the baseline current
#   event_current: (float) event current
#   event_sd: (float) square root of the variance of the event current
#   residual_current: (float) residual current fraction of the event
#   residual_current_sd: (float) square root of the variance of the residual current (error propagated)
#   dwell_time: (float) residence time of the event in seconds
features = event_analysis.features

The events and features can be accessed using named tuples, which can be used to, for example, calculate the excluded current.

excluded_current = np.array([1-i.residual_current for i in features]) # excluded current fraction
excluded_current_sd = np.array([i.residual_current_sd for i in features]) # The square root of the variance is equal to the residual

For the unbiased data analysis for the parameterization of fast translocation events run the following code. This will fit a function (default generalised Normal Distribution Function) around each event.

results = event_analysis.optimise_events()

# returns a list of named tuples, each element in the list is an optimized event.
# - Method: (str) method used for optimisation
# - Trace: (int) trace number
# - Function: (str) function used for optimisation, e.g. gNDF
# - Fitting_parameters: (tuple) parameters for Function
# - Amplitude_block: (float) amplitude of the block
# - Localisation: (float) centroid of the block in time
# - Sigma: (float) sigma over the localisation
# - Beta: (float) beta parameter
# - Open_current: (float) open pore current
# - Dwell_time: (float) dwell time of the event
# - Fs_event: (float) effective frequency

Authors and acknowledgements

Main contributors

• Florian L. R. Lucas
• Matthijs Tadema

Acknowledgements

• Maglia lab, University of Groningen

Changelog

12-10-2022: Updated readme page

How to cite

When using this library, please be kind to cite the following work:

Unbiased Data Analysis for the Parameterization of Fast Translocation Events through Nanopores Florian L. R. Lucas, Kherim Willems, Matthijs J. Tadema, Katarzyna M. Tych, Giovanni Maglia, and Carsten Wloka ACS Omega 2022 7 (30), 26040-26046 DOI: 10.1021/acsomega.2c00871