These scripts demonstrate the use of bayesian metamodeling of complex biological systems across varying representations.

Authors: Liping Sun, Barak Raveh

License:

Publications:

Prerequisites:

• matlab: The scripts are built upon and work with Matlab.
• bnt: Bayes Net Toolbox for Matlab.

To get started:

• Either use the bnt version here (under bnt_master/), or download a recent version of bnt here: https://github.com/bayesnet/bnt.

• If downloaded, apply two minor bugfixes on bnt:

  • Disable line 132 of /Your-Path-To/bnt-master/BNT/learning/learn_params_dbn_em.m.

  125- loglik = 0;
  126- for l=1:length(cases)
  127-   evidence = cases{l};
  128-   if ~iscell(evidence)
  129-     error('training data must be a cell array of cell arrays')
  130-   end
  131-   [engine, ll] = enter_evidence(engine, evidence);
  132-   % assert(~isnan(ll))
  133-   loglik = loglik + ll;
  134-   T = size(evidence, 2);

  • Wrap line 85 with the following if statement in /Your-Path-To/bnt-master/BNT/general/mk_bnet.m.

  85-  if length(mems)>=1
  86-    bnet.rep_of_eclass(e) = mems(1);
  87-  end

• Add the path of the bnt package:

  1- addpath(genpathKPM('/Your-Path-To/bnt-master'))

List of files and directories:

• data contains the data of six input models and the metamodel including:
  • JSON files with the values of model parameters and variables:
  • GI.dat with the observed values for the glucose intake after a meal
  • Gb_kt_input_err101.dat and Gb_kt_input_sigma101.dat with the input values for different accuracy and precision of model variables G_B and kt
  • 072919-INS1e-30min-Enrichment-analysis-cleaned-summary.xlsx with the data for the metabolism model
• scripts contains all the bnet scripts for metamodeling, please refer to README.md in scripts for more details
• bnt_master contains Bayes Net Toolbox for Matlab with the bugfixes.