cluster
This module provides the ProcessProtein class, which collects and processes Gibbs sampler data.
- class cluster.ProcessProtein(niter, prot, cutoff, gskip=100, burnin=10000, taus=None, bars=None)[source]
ProcessProtein is the class that collects and processes Gibbs sampler data. This class collects results for all residues in the basicrta-{cutoff} directory and can write out a \(\tau\) vs resid numpy array or plot \(\tau\) vs resid. If a structure is provided, \(\tau\) will be written as b-factors for visualization.
- Parameters:
niter (int) – Number of iterations used in the Gibbs samplers
prot (str, optional) – Name of protein in tm_dict.txt, used to draw TM bars in \(tau\) vs resid plot.
cutoff (float) – Cutoff used in contact analysis.
gskip (int) – Gibbs skip parameter for decorrelated samples; only save every gskip samples from full Gibbs sampler chain; default from https://pubs.acs.org/doi/10.1021/acs.jctc.4c01522 When the sampled Markov chain is loaded, then the output is already saved at every Gibbs.g samples. We calculate a new gskip value to get close to the desired gskip value.
burnin (int) – Burn-in parameter, drop first burnin samples as equilibration; default from https://pubs.acs.org/doi/10.1021/acs.jctc.4c01522
- b_color_structure(structure)[source]
Add \(\tau\) to b-factors in the specified structure. Saves structure with b-factors to tau_bcolored.pdb.
- get_taus(nproc=1)[source]
Get \(\tau\) and 95% confidence interval bounds for the slowest process for each residue.
- Returns:
Returns a tuple of the form (\(\tau\), [CI lower bound, CI upper bound])
- Return type:
- plot_protein(**kwargs)[source]
Plot \(\tau\) vs resid. kwargs are passed to the
plot_protein()method of util.py. These can be used to change the labeling cutoff, y-limit of the plot, scale the figure, and set major and minor ticks.