Modeling Signal Transduction with Boolean Paradigms M. ARMAS, O. ARMAS, M. CÁRDENAS, J. VASCONCELOS, & J. LAGÚNEZ-OTERO Instituto de Química Universidad Nacional Autónoma de México México, D.F., 04510, MEXICO Abstract Proteins and their interactions are the basis for the interpretation of DNA into a phenotypic trait. In order to aid in understanding this process, the relevant knowledge may be structured in computational formulations. We are currently developing a knowledge-based system (GENIA) to study the relationship between protein interactions, expression and phenotypic manifestations. Though the emphasis of the model is on applications to gene therapeutics, it is also a tool for a general molecular biologist. We are currently including the rules for the modeling of genetic regulatory cycles of oncogenic origin. Keywords: Bioinformatics, proteomics, logic programming, knowledge base systems, expert systems 1 Introduction The determination of the structure of DNA and the description of genes as nucleic acid sequences, allowed for the first stages for growth in the areas of molecular biology and genetic engineering. Later, the finding of phosphorylation cascades within signal transduction systems generated exciting breakthroughs in the comprehension of cell development and pathologies such as cancer. These two aspects of management correspond to storage and transmission of information. Information in the cell however goes beyond these. Presently, we are at the point of cracking the code of information evaluation and the decision making by the cell, believing that it will also yield important discoveries. We see that the problems of information management, especially evaluation and decision can best be dealt with using boolean models. in particular, logic programming and expert systems. We are developing a system representing the logic of signal trasduction. It includes the phosphorylation cascade corresponding to the transmission of the epidermal growth factor signal which uses RAS or MEK - SEK. We are able to model this transmission by including the if-then rules for the presence of the EGF, its receptor and the following events leading to the initiation of transcription. We have found that the protein 14-3-3 could be an ideal candidate for inhibition in order to reach a necrotic process. This protein has been shown to accelerate the processes in which it is involved and most important, there are no routes that substitute its activity. The system shell allows for object oriented programing as well as If- Then production rules. 2 Object Oriented Programing Objects are organized according to the classes they belong to: E.g.: Protein -> enzyme -> G-Protein -> Ras family -> Hras their properties, which can be inherited, are concentration, state of activity, size and phosphorylation sites. (Figures 1a and 1b) Rules are of the type if-then-else. Mutations are assigned before the inference process is begun by giving a value of 0 to concentration. Fig 1a. Objects in the data base