Identifying DNA splice sites using hypernetworks with artificial molecular evolution.

Identifying DNA splice sites is a main task of gene hunting. We introduce the hyper-network architecture as a novel method for finding DNA splice sites. The hypernetwork architecture is a biologically inspired information processing system composed of networks of molecules forming cells, and a number of cells forming a tissue or organism. Its learning is based on molecular evolution. DNA examples taken from GenBank were translated into binary strings and fed into a hypernetwork for training. We performed experiments to explore the generalization performance of hypernetwork learning in this data set by two-fold cross validation. The hypernetwork generalization performance was comparable to well known classification algorithms. With the best hypernetwork obtained, including local information and heuristic rules, we built a system (HyperExon) to obtain splice site candidates. The HyperExon system outperformed leading splice recognition systems in the list of sequences tested.

The effect of molecular inhibition on evolutionary learning: studies in the hypernetwork architecture.

The hypernetwork architecture is a biologically inspired learning model based on abstract molecules and molecular interactions that exhibits functional and organizational correlation with biological systems. Hypernetwork organisms were trained, by molecular evolution, to solve N-input parity tasks. We found that learning improves when molecules exhibit inhibitory sites, allowing molecular inhibition and opening the possibility of forming negative feedback regulatory pathways. Optimal learning is achieved when at least 20% of the molecules in each cell have inhibitory sites. Intra-cellular as well as inter-cellular molecular inhibitions play an important role in the information processing of hypernetwork organisms, by maintaining a balance of the molecular cascade reactions. Similar mechanisms inside neurons are considered important for memory.