Modeling dynamics of FS Peptide

This example shows how to use sklearn's Pipeline object to rapidly construct MSMs

MSMBuilder includes example datasets

The following cell loads an example protein dataset. In practice, you will generate your own dataset by running molecular dynamics (MD) on your system of interest. MSMBuilder does not run MD.



In [ ]:

    
# Download example dataset
from msmbuilder.example_datasets import FsPeptide
fs_peptide = FsPeptide()
xyz = fs_peptide.get().trajectories
print(fs_peptide.description())

Since the data was saved at 50 ps / frame, we only load every 10th frame (with a new frequency of 0.5/ns).



In [ ]:

    
xyz = [t[::10] for t in xyz]
print("{} trajectories".format(len(xyz)))
# msmbuilder does not keep track of units! You must keep track of your
# data's timestep
to_ns = 0.5
print("with length {} ns".format(set(len(x)*to_ns for x in xyz)))

Pipeline



In [ ]:

    
from sklearn.pipeline import Pipeline
from msmbuilder.featurizer import DihedralFeaturizer
from msmbuilder.preprocessing import RobustScaler
from msmbuilder.decomposition import tICA
from msmbuilder.cluster import MiniBatchKMeans
from msmbuilder.msm import MarkovStateModel

pipeline = Pipeline([
        ('feat', DihedralFeaturizer()),
        ('scaler', RobustScaler()),
        ('tica', tICA()),
        ('cluster', MiniBatchKMeans(n_clusters=100, random_state=42)),
        ('msm', MarkovStateModel()),
    ])



In [ ]:

    
pipeline = pipeline.fit(xyz)

Use our MSM



In [ ]:

    
msm = pipeline.named_steps['msm']



In [ ]:

    
%matplotlib inline
from matplotlib import pyplot as plt

plt.scatter(msm.left_eigenvectors_[:,1],
            msm.left_eigenvectors_[:,2],
            s=msm.populations_*1e4,
            c=msm.left_eigenvectors_[:,1],
            cmap='coolwarm')
plt.tight_layout()



In [ ]:

    
import msmexplorer as msme
msme.plot_timescales(msm, n_timescales=5, ylabel='Implied Timescales ($ns$)')