Practical Inference Methods for Mechanistic Modelling of Biological Systems

Bayesian Inference of Mechanistic Systems Models Using Population MCMC

author: Ben Calderhead, University of Glasgow

Description

We demonstrate how Population Markov Chain Monte Carlo techniques may be used to sample from the complex posterior distributions which arise when estimating parameters over nonlinear mechanistic mathematical models of biological processes given noisy data. Further, we show how the samples obtained may be employed, using a Power Posteriors method, to accurately calculate the marginal likelihoods and Bayes factors over such models.

You might be experiencing some problems with Your Video player.

*Slides*
0:00	Bayesian Inference of Mechanistic System Models Using Population MCMC
0:54	Outline
1:17	Oscillatory Control Processes - 1
1:29	Oscillatory Control Processes - 2
2:12	Oscillatory Control Processes - 3
2:29	Oscillatory Control Processes - 4
2:38	The Challenge of Identifiability - 1
3:09	The Challenge of Identifiability - 2
3:36	The Challenge of Identifiability - 3
4:21	A Multi-Dimensional Representation of a Complex Posterior Distribution - 1
4:48	A Multi-Dimensional Representation of a Complex Posterior Distribution - 2
5:54	Parameter Inference
5:57	A Minimal Model to Describe Circadian Networks
6:34	Goodwin Model - 1
6:47	Goodwin Model - 2
6:56	Goodwin Model - 3
7:37	- Questions
10:12	Population Markov Chain Monte Carlo - 1
10:56	Population Markov Chain Monte Carlo - 2
11:24	Population Markov Chain Monte Carlo - 3
12:18	Population MCMC Algorithm
14:15	- Questions
15:41	Exploring a Goodwin Model - 2
16:45	Model Comparison
16:57	Bayes Factors - 1
17:08	Bayes Factors - 2
17:46	Bayes Factors - 3
17:52	Thermodynamic Integration - 1
18:34	Thermodynamic Integration - 2
18:52	Discretised Thermodynamic Integral - 1
19:02	- Questions
19:57	Optimal Temperature Schedules - 1
20:17	Optimal Temperature Schedules - 2
20:33	Optimal Temperature Schedules for Linear Model
21:38	Goodwin Model Identification - 1
21:58	Goodwin Model Identification - 2
22:27	Goodwin Model Identification - 3
23:09	Results - Metropolis-Based Sampling - 1
23:37	- Questions
24:52	Results - Population MCMC - 1
25:59	Results - Population MCMC - 2
26:04	Conclusions - 1
26:14	Conclusions - 2
26:38	- Questions

Lecture rating

People found this lecture:
Worth seeing
because it is:
Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.