Abstract—Cytocomputation is a computational paradigm based upon the macromolecular activity inside the cytoplasm of the biological cells. This paradigm can be used either as a source of inspiration for proposing novel computational architectures, or as a framework for modeling biological processes at the intracellular and intercellular levels. This paper presents the main characteristics of the paradigm and describes its implementation on the ubichip, a hardware platform specifically designed to support bioinspired architectures. I. INTRODUCTION ATURE has been used as a source of inspiration by engineers and computer scientists since several decades ago. Conspicuous examples of this tendency are the artificial versions of neural networks, ant colonies, genetic algorithms, or immune systems. In all these cases, their designers have turned their attention to particular aspects of Nature, extracted a set of representative features, proposed a simplified parameter-based model, and applied it to specific problems [1]. One remarkable characteristic of the bioinspired systems is that they do not share common features, but the name. Consequently, it is difficult to integrate them smoothly into a unified system, because of the lack of a common base [2]. Cytocomputation [3] is a proposal that attemps to provide a unified framework for bioinspired computation. To do this, Nature has been revisited and the fundamental biological entity, the cell, has been taken as the main metaphor. Cytocomputation is based on the macromolecular activity inside the cytoplasm of the biological cells. The model is composed of four computational spaces: protein, enzyme, genetic, and phenotypic spaces. [4] Protein computation follows the metaphor of a folding protein, where relative bindings are modified during the process of solving a given problem. Enzyme computation is Manuscript received November 13, 2008. This work was supported in part by Colciencias, the Colombian Science Foundation. J. A. Parra-Plaza is with the Computer Science and Engineering Department, Pontificia Universidad Javeriana at Cali, Colombia and is currently a doctorate student at the Universidad del Valle, Cali, Colombia (e- mail: jparra@javerianacali.edu.co). Andres Upegui is a Senior Researcher at the Institute of Reconfigurable and Embedded Digital Systems (REDS), at the University of Applied Sciences of Western Switzerland (HEIG-VD), Rte de Cheseaux 1, 1400 Yverdon-les- Bains, (email: andres.upegui@heig-vd.ch).. J. Velasco is with the School of Electrical and Electronics Engineering, Universidad del Valle, Cali and is the director of the Bionanoelectronics group (e-mail: jvelasco@univalle.edu.co). performed through rewriting rules, where collisions between proteins occur in a random way and as a result they are modified according to the currently active enzymes. Genetic computation is performed by a set of genes producing proteins, enzymes, and transcription factors. Proteins holds data, enzymes transform proteins, and transcription factors regulate other genes. Phenotypic computation allows evolving the products generated for the other spaces, differentiating between the search space (genotype) and the solutions space (phenotype). Cytoelectronics [5] is the hardware version of the cytocomputational paradigm and, in its current state, is composed of the protein and genetic spaces. It has been developed in Perplexus [6], a hardware platform conceived to ease the implementation of bioinspired circuits. Two remarkable features that contribute to this effort are dynamic routing [7] and dynamic reconfiguration [8]. Both nicely fit with the cytocomputational model. This paper is organized as follows: section II gives the biological support. Section III presents the computational model. Section IV explains the hardware implementation. Section V offers the conclusions and future work. II. BIOLOGICAL BACKGROUND Every complex organism on Earth is built of cells. A cell is both a building block for the entire organism and an individual entity capable of managing its own development and maintenance. Multi-cellular organisms are comprised of eukaryotic cells, thus they have an identifiable nucleus. Structurally, it is possible to recognize different compartments into a cell, each one surrounded by a membrane, including the cell itself. The most remarkable among them is the nucleus, which houses the genome, coded into a linear sequence of DNA. The features of the genome, or genotype, are expressed through the creation of proteins, which constitute the components for structure and function of the cell, or phenotype (see Figure 1), through protein associations, post-translational modifications, membrane insertion, or intercellular signaling. From a computational point of view, the most important molecules in a cell are those called macromolecules [9], which are mainly the DNA, the proteins, and the enzymes. Although, structurally, an enzyme is a protein; functionally it is very different. In fact, the task of an enzyme is usually to modify the structure of a protein [10]. Cytocomputation in a biologically inspired, dynamically reconfigurable hardware platform J. A. Parra-Plaza, A. Upegui, and J. Velasco-Medina, IEEE Senior Member N