Proceedings of CIBB 2011 1 Modelling the Effect of Genes on the Dynamics of Probabilistic Spiking Neural Networks for Computational Neurogenetic Modelling Nikola Kasabov 1,2 , Stefan Schliebs 1 , Ammar Mohemmed 1 1 Knowledge Engineering and Discovery Research Institute Auckland University of Technology, Auckland 1010, New Zealand {nkasabov, sschlieb,amohemme}@aut.ac.nz 2 Institute for Neuroinformatics, University of Zurich and ETH Keywords: Computational neurogenetic modeling; Spiking neural networks; Gene reg- ulatory networks; Probabilistic neural models. Abstract. Computational neuro-genetic models (CNGM) combine two dynamic mod- els – a gene regulatory network (GRN) model at a lower level, and a spiking neural network (SNN) model at a higher level to model the dynamic interaction between genes and spiking patterns of activity under certain conditions. The paper demonstrates that it is possible to model and trace over time the effect of a gene on the total spiking behavior of the SNN when the gene controls a parameter of a stochastic spiking neuron model used to build the SNN. Such CNGM can be potentially used to study neurodegenerative diseases or develop CNGM for cognitive robotics. 1 Introduction Computational Neuro Genetic Modelling (CNGM) is a biologically motivated mod- eling approach that is concerned with the creation of two-level hierarchical computa- tional models, where interaction between large number of dynamic variables (called genes) is modeled over time as a gene-regulatory network model (GRN) that affects the activity of a higher level system – modeled as a spiking neural network (SNN). The behavior of the two systems, in their continuous interaction under certain input- output conditions, have been introduced and studied in [Kasabov and Benuskova, 2004, Kasabov and Benuskova, 2005, Kasabov et al., 2005, Kasabov, 2007, Kasabov, 2009, Kasabov, 2008, Kasabov et al., 2011]. CNGM constitute the next generation of compu- tational modeling techniques built on the foundations of the traditional neural network techniques. The goal of this paper is to explore and to develop further the CNGM paradigm through the introduction of stochastic neuronal models (e.g. [Gerstner and Kistler, 2002]) used to build stochastic/probabilistic SNN (pSNN). Such pSNN are more biologically plausible, offering some additional advantages [Kasabov et al., 2011]. For this purpose genes are used to control parameters and their effect on the behavior of the whole pSNN is modeled and studied. A specific gene from the genome relates to the activity of a neuronal cell by means of a specific protein. Complex interactions between genes and proteins within the in- ternal gene/protein regulatory network influence the functioning of each neuron and a neural network as a whole [Marcus, 2005, Holter et al., 2001]. With the advancement of molecular research technologies huge amount of data and information is available about the genetic basis of neuronal functions and diseases [Marcus, 2005, Holter et al., 2001, NCBI, 2003]. This information can be utilized to create models of brain functions and diseases that include models of gene interactions within models of neural networks. In order to create biologically plausible CNGM we need to integrate knowledge from genomics, proteomics, neuroscience, psychology, and theoretical disciplines such as