Aberrant gene functions usually contribute to the pathology or diseases. Avoiding undesirable cellular phenotypes as many as possible is a major purpose of external control for gene regulatory networks. An interesting question is how to control a gene network subjected to the condition that the genes reach some undesirable states with minimal probability during a cell cycle. In this paper, we make use of the theory of the first passage model for discrete-time Markov decision processes to determine the optimal control for a gene intervention model. Specifically, we first use a control model for a probabilistic Boolean network to model interactions among genes and then solve an optimal control problem for maximising the probability of the first arrival time to desirable gene states. In order to illustrate the validity of our approach, examples are also displayed.
[1]
D. Gillespie.
Exact Stochastic Simulation of Coupled Chemical Reactions
,
1977
.
[2]
Stuart A. Kauffman,et al.
The origins of order
,
1993
.
[3]
Stuart A. Kauffman,et al.
ORIGINS OF ORDER IN EVOLUTION: SELF-ORGANIZATION AND SELECTION
,
1992
.
[4]
Stella X. Yu,et al.
Optimization Models for the First Arrival Target Distribution Function in Discrete Time
,
1998
.
[5]
Siming Huang,et al.
Markov Decision Processes with Distribution Function Criterion of First-Passage Time
,
2001
.
[6]
Aniruddha Datta,et al.
External Control in Markovian Genetic Regulatory Networks
,
2004,
Machine Learning.
[7]
Xianping Guo,et al.
Optimal risk probability for first passage models in semi-Markov decision processes
,
2009
.