PyBroMo - 2. Generate smFRET data, including mixtures

This notebook is part of PyBroMo a python-based single-molecule Brownian motion diffusion simulator that simulates confocal smFRET experiments.

Overview

In this notebook we show how to generated smFRET data files from the diffusion trajectories.

Loading the software

Import all the relevant libraries:



In [ ]:

    
%matplotlib inline
from pathlib import Path
import numpy as np
import tables
import matplotlib.pyplot as plt
import seaborn as sns
import pybromo as pbm
print('Numpy version:', np.__version__)
print('PyTables version:', tables.__version__)
print('PyBroMo version:', pbm.__version__)

Create smFRET data-files

Create a file for a single FRET efficiency

In this section we show how to save a single smFRET data file. In the next section we will perform the same steps in a loop to generate a sequence of smFRET data files.

Here we load a diffusion simulation opening a file to save timstamps in write mode. Use 'a' (i.e. append) to keep previously simulated timestamps for the given diffusion.



In [ ]:

    
S = pbm.ParticlesSimulation.from_datafile('0168', mode='w')



In [ ]:

    
S.particles.diffusion_coeff_counts



In [ ]:

    
#S = pbm.ParticlesSimulation.from_datafile('0168')

Simulate timestamps of smFRET

Example1: single FRET population

Define the simulation parameters with the following syntax:



In [ ]:

    
params = dict(
    em_rates = (200e3,),    # Peak emission rates (cps) for each population (D+A)
    E_values = (0.75,),     # FRET efficiency for each population
    num_particles = (20,),  # Number of particles in each population
    bg_rate_d = 1500,       # Poisson background rate (cps) Donor channel
    bg_rate_a = 800,        # Poisson background rate (cps) Acceptor channel
    )

Create the object that will run the simulation and print a summary:



In [ ]:

    
mix_sim = pbm.TimestapSimulation(S, **params)
mix_sim.summarize()

Run the simualtion:



In [ ]:

    
rs = np.random.RandomState(1234)
mix_sim.run(rs=rs, overwrite=False, skip_existing=True)

Save simulation to a smFRET Photon-HDF5 file:



In [ ]:

    
mix_sim.save_photon_hdf5(identity=dict(author='John Doe', 
                                       author_affiliation='Planet Mars'))

Example 2: 2 FRET populations

To simulate 2 population we just define the parameters with one value per population, except for the Poisson background rate that is a single value for each channel.



In [ ]:

    
params = dict(
    em_rates = (200e3, 180e3),   # Peak emission rates (cps) for each population (D+A)
    E_values = (0.75, 0.35),     # FRET efficiency for each population
    num_particles = (20, 15),    # Number of particles in each population
    bg_rate_d = 1500,       # Poisson background rate (cps) Donor channel
    bg_rate_a = 800,        # Poisson background rate (cps) Acceptor channel
    )



In [ ]:

    
mix_sim = pbm.TimestapSimulation(S, **params)
mix_sim.summarize()



In [ ]:

    
rs = np.random.RandomState(1234)
mix_sim.run(rs=rs, overwrite=False, skip_existing=True)



In [ ]:

    
mix_sim.save_photon_hdf5()

Burst analysis

The generated Photon-HDF5 files can be analyzed by any smFRET burst analysis program. Here we show an example using the opensource FRETBursts program:



In [ ]:

    
import fretbursts as fb



In [ ]:

    
filepath = list(Path('./').glob('smFRET_*'))
filepath



In [ ]:

    
d = fb.loader.photon_hdf5(str(filepath[1]))



In [ ]:

    
d



In [ ]:

    
d.A_em



In [ ]:

    
fb.dplot(d, fb.timetrace);



In [ ]:

    
d.calc_bg(fun=fb.bg.exp_fit, tail_min_us='auto', F_bg=1.7)



In [ ]:

    
d.bg_dd, d.bg_ad



In [ ]:

    
d.burst_search(F=7)



In [ ]:

    
d.num_bursts



In [ ]:

    
ds = d.select_bursts(fb.select_bursts.size, th1=20)



In [ ]:

    
ds.num_bursts



In [ ]:

    
fb.dplot(d, fb.timetrace, bursts=True);



In [ ]:

    
fb.dplot(ds, fb.hist_fret, pdf=False)
plt.axvline(0.75);

NOTE: Unless you simulated a diffusion of 30s or more the previous histogram will be very poor.



In [ ]:

    
fb.bext.burst_data(ds)



In [ ]: