pysmoke 0.1.0

Small Molecule Octet/BLI Kinetics Experiment

smoke-bli - Small Molecule Octet/BLI Kinetics Experiment

A Python package for processing and analyzing Bio-Layer Interferometry (BLI) kinetics data acquired on Octet Red96 (Fortebio).

BLI signals are small and noisy when small molecules are investigated as ligands (analytes). This package provides tools to accurately process and fit 1:1 binding kinetics from these experiments.

Reference: Sun Q., Li X., et al (2020). Nature Communications, 11: 339. doi:10.1038/s41467-019-14238-3

Converted from the original R/CRAN package smoke (v2.0.1) by Qingan Sun, Xiaojun Li, and James C Sacchettini.

Installation

pip install pysmoke

Dependencies

• numpy
• pandas
• scipy
• matplotlib
• corner

Quick Start

import pandas as pd
from smoke import Bli

# Create a Bli object
bli = Bli()

# Load your trace data (first column = time, remaining columns = sensor traces)
bli.traces = pd.read_csv("my_traces.csv", header=None)

# Set ligand concentrations (high to low) and experiment times
bli.ligand = [16.0, 8.0, 4.0, 2.0, 1.0, 0.5, 0.25]
bli.t_exp = [1260, 1860]  # [association_start, dissociation_start]

# Processing pipeline
bli.align_load(load_start=180, load_end=780)
bli.double_blank()
bli.baseline_correct(t_start=1080, t_end=1260)

# Estimate initial rates and fit global kinetics
bli.estimate()

# Option 1: Analytical kinetics fitting
bli.fit_kinetics()

# Option 2: ODE-based kinetics fitting
bli.ode_kinetics(negative_response=False)

# View results
print(bli.get_kinetics())

# Plot
bli.plot_traces()
bli.plot_kinetics()
bli.plot_residuals()
bli.plot_confidence_contours()

API Reference

Bli Class

Attributes

Attribute	Type	Description
traces	pd.DataFrame	Trace data. First column is time, remaining columns are sensor traces.
ligand	np.ndarray	Ligand concentrations, ordered high to low.
t_exp	np.ndarray	Two-element array: [association_start, dissociation_start].
k_on0	float or None	Initial on-rate estimate.
k_off0	float or None	Initial off-rate estimate.
status	dict	Processing step completion flags.
kinetics	dict	Fitted model results (populated after fit_kinetics() or ode_kinetics()).

Processing Methods

Method	Description
align_load(load_start, load_end)	Align paired traces to the loading step.
double_blank()	Apply double-blank (double-reference) correction.
baseline_correct(t_start, t_end)	Subtract a linear baseline fitted over a time window.
estimate()	Estimate initial k_on0 and k_off0 from individual trace fits.
fit_kinetics(negative_response=False)	Global fit of 1:1 binding kinetics with linear drift correction. Set negative_response=True for decreasing signals.
ode_kinetics(negative_response=True)	ODE-based global fit of 1:1 binding kinetics with linear drift correction.

Result Methods

Method	Description
get_kinetics()	Returns a DataFrame with KD, rMax, kOn, kOff (estimate, std error, t-value, p-value).

Plotting Methods

Method	Description
plot_traces(**kwargs)	Plot all traces with cyan-to-magenta gradient.
plot_kinetics(**kwargs)	Plot traces with fitted kinetics overlay.
plot_residuals(**kwargs)	Plot residuals from the global fit.
plot_confidence_contours()	Corner plot showing confidence contours for kOn, kOff, and rMax.

R to Python Mapping

R (original)	Python
Bli()	Bli()
traces(obj) / traces(obj)<-	obj.traces
ligand(obj) / ligand(obj)<-	obj.ligand
tExp(obj) / tExp(obj)<-	obj.t_exp
kOn0(obj) / kOn0(obj)<-	obj.k_on0
kOff0(obj) / kOff0(obj)<-	obj.k_off0
status(obj)	obj.status
alignLoad(obj, loadStart, loadEnd)	obj.align_load(load_start, load_end)
doubleBlank(obj)	obj.double_blank()
baseline(obj, tStart, tEnd)	obj.baseline_correct(t_start, t_end)
estimate(obj)	obj.estimate()
fitKinetics(obj)	obj.fit_kinetics()
odeKinetics(obj)	obj.ode_kinetics()
kinetics(obj)	obj.get_kinetics()
plotTraces(obj)	obj.plot_traces()
plotKinetics(obj)	obj.plot_kinetics()
plotResiduals(obj)	obj.plot_residuals()
plotConfidenceContours(obj)	obj.plot_confidence_contours()

License

GPL-3.0-or-later