Modeling Genetic Regulatory Networks Using First-Order Probabilistic Logic

Abstract

New technologies such as microarrays and flow cytometry have led to the availability of large amounts of biological data. There is a need to model biological systems to aid in medication and drug delivery. Genetic Regulatory Networks (GRNs) represent the signal transduction, or the activation and deactivation of genes, as their corresponding proteins directly or indirectly interact with one another. GRNs can be modeled using statistical and logical techniques, more precisely using Bayesian networks. Bayesian networks are directed acyclic graphs (DAGs) where the nodes represent random variables and edges represent conditional dependencies. In this research, a learning algorithm was implemented to determine the structure and the parameters of Bayesian networks that model GRNs from real data. PRISM, a probabilistic learning framework based on B-prolog, was used to program the Bayesian networks. Instead of conventional statistical techniques, which rely on point estimates, the method of Variational Bayes-Expectation Maximization (VB-EM) was used to obtain a lower-bound to the marginal likelihood, or the free energy, and a set of optimal parameters. A hill-climbing algorithm using this free energy as a scoring function was utilized. The learning algorithm was tested on the well-established Raf pathway.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 2013
Accession Number
ADA582376

Entities

People

  • Nand Kishore
  • Radhakrishnan Balu
  • Shashi P. Karna

Organizations

  • United States Army Research Laboratory

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Algorithms
  • Artificial Intelligence
  • Bayesian Networks
  • Cell Physiological Processes
  • Chemistry
  • Computational Science
  • Computer Programming
  • Computer Programs
  • Diseases And Disorders
  • Free Energy
  • Machine Learning
  • Models
  • Probabilistic Models
  • Probability
  • Probability Distributions
  • Random Variables
  • Systems Biology

Readers

  • Graph Algorithms and Convex Optimization.
  • Molecular Genetics
  • Statistical inference.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms
  • AI & ML - Neural Networks
  • Biotechnology