Journal of Computational Intelligence in Bioinformatics
  • Year: 2010
  • Volume: 3
  • Issue: 1

Markov Properties of Graph Models and Probabilistic Inference on Bayesian Networks in Genetics

  • Author:
  • K.A. Germina1, M. Kumaran2, Geetha Antony Pullen3
  • Total Page Count: 16
  • Page Number: 111 to 126

1PG & Research Department of Mathematics, Mary Matha Arts & Science College, Mananthavady

2Department of Statistical Sciences, Kannur University, Kannur

3Mary Matha Arts & Science College, Vemom P.O., Mananthavady -670645, Kerala, India. E-mail: geethapullen@gmail.com

Abstract

Probability propagation in trees of clusters (PPTC) is an algorithm that provides an easy method for probability inference. It uses the markov properties of graph models, and the underlying graph model is converted into a secondary structure, on which probabilities are calculated. In the present study, the authors illustrate PPTC method by computing probabilities using transition probability matrix (tpm) and using the PPTC algorithm on a simple genetic inheritance model and the results are compared.

Keywords

Graph theory, digraphs, recursive factorization of probability on graph models, markov properties, PPTC method