Algebraic& Statistic Computation for Model Checking in BioInformatics

Speaker: Quoc-Nam Tran
Date: 04/06/2012
Time: 13:30 - 14:30

Location: RISC Seminar room

Temporal model checking is an algorithmic and formal approach for automatically verifying whether a finite-state concurrent system such as a sequential circuit design functions correctly. Typically, computation is carried over Boolean algebras using binary decision diagrams (BDDs) or satisfiability (SAT) solvers. Researchers have been using BDDs and bounded model checking (BMC) on Boolean gene regulatory networks for bioinformatics. Previous works also showed that BDDs blow-up more frequently on random networks than on sequential circuits. A gene regulatory network is a collection of DNA segments in a cell which interact with each other indirectly through their RNA and protein expression products and with other substances in the cell, thereby governing the rates at which genes in the network are transcribed into mRNA. We present a computational method for direct computation of Groebner bases (GB) in Boolean rings for temporal logic reasoning and for checking the dynamic of Boolean gene regulatory networks in particular. In contrast to other known algebraic approaches, the degree of intermediate polynomials during the calculation of Groebner bases using our method will never grow resulted in a significant improvement in running time and memory space consumption. Another focus of this talk is on probabilistic boolean networks because recent experimental results have demonstrated that gene expression is a stochastic process. We will show how algebraic and continuous stochastic logic can be used for checking the dynamic of probabilistic boolean networks in the context of Markov theory.