A software package for quantifying information accumulation of biochemical signaling channels.
(c) 2020 Signaling Systems Lab, UCLA All rights reserved. This MATLAB code package implements the quantification on the dynamical mutual information for data of time series. It used either time-inhomogeneous Markov model or hidden Markov model to learn the dynamical patterns of the set of time series. Then, it can reproduce the path ensemble by sampling a same amount of time series, and quantify the similarity between data and sampling. It further calculates the trajectory probability for each time series, and the dynamical mutual information.
A detailed example on the methods is given in the main text.
A guideline for the package is on the website: the website link will be generated after the manuscript is accepted.
Contact: Ying Tang, jamestang23@gmail.com
Guideline
A step-by-step guideline is on the website of the package: https://sites.google.com/view/dmipackage. The website link may differ after the manuscript is accepted.
System requirements: All simulations were done using MATLAB® version R2019a. We have used the toolbox “Hidden Markov Models (HMM)” in MATLAB.
Third-party packages:
(1) We have used the package of NFkB signaling model on https://github.com/biomystery/tnf_ikbd_nfkb_model.git.
(2) The package to generate the data of NFkB on https://github.com/Adewunmi91/MACKtrack.
(3) We thank Roy Wollman's group for sharing the code of vector method.
(4) The decoding-based method was not included here, because it can be separately implemented by the user-friendly package (https://github.com/swainlab/mi-by-decoding).
Expected run time: all the expected run time below is evaluated based on a personal desktop with intel(R) core(tm) i7-8700 CPU @ 3.7GHz.