BioSystems 39 (1996) 143-151 Self-organization in computer simulated selective systems Sergei P. Atamas Division of Rheumatology and Clinical lmmunology, Department of híedicine, School of Medicine, University of Maryhmd at Baltimore, 10 S. Pine St., MSTF. 8-22, Baltimore, MD 21201, USA Received 23 October 1995; revised 29 January 1996; accepted 12 February 1996 AbEhCt In certain biologie systems, the signal selects functional or numerical expansion of the recognizing element& Ex- amples of these systems include the immune system, brain cortex, and evolution. One common feature of these Danvinian-type systems is degenerate recognition, in which one signal can recognize several different elements, with difkent aRnities and consequences. For example, T cel1 antigen receptors and antibodies demonstrate relative but not absolute specifmity of recognition. Thus, the variables of dose of the signal and affinity of the recognixing element modulate the outcome. Another feature of these systems is the ability to create self-organixed pattems, which do not mirror the incoming signals. The hypothesis of this study is that degenerate recognition with subsequent selection of recognixing ekments can explain self-organixation of these systems. An entirelymunerical model was explored, using the cellularautomata approach.Three intrinsic featuresof a common selectivesystemwere incorporated into this model:a largenumber of recoguixing elements; degenerative recognitionof stimuliby these elements; and subsequent selection. Ditkrent numerical pattems of incoming stimuli were tested. The model showed self-organixing dynamics. Usually, the population of recognixing elements demonstrated an initial period of equilibrium, then a chaotic transi- tional state, and, fínally, the bifurcational appearance of a stable self-organixed pattem. The tinal resolution into a stable pattem can be either gradual or quasi-saltational. 1 conclude that systems with a large number of recognixing elements, degenerative recognition, and selection of recognixing elements can self organixe based upon the pattem of the incoming stimuli. Keywordr: Selective system; Self-organixation; Degenerate recognition; Competition; Computer modeling 1. -00 Inclusion of Darwinism in a broader context, such as a theory of selI+anization, seems to be one of the main trends in modem theoretical bio- logy (Nicolis and Prigogine, 1977; Babloyantz, ?? Correqonding author. Tel.: +1410 706320% Fax: +l 410 7060231; E-mail:Wamas~bnet.ab.~.edu 1986; KauEnan, 1993). The last decades have in- troduced several new terms and entirely new ap- proaches into contemporary biological Wiking. Among them are deterministic chaos, fractals, ex- citable cel1 assemblies, cellular automata, etc. Bio- logica1 development in the broad sense appears now as a much more complicated process than Darwinian selection. Amazing self-organizing pro- cesae~ take place at al1 levels of biology from the 0303-2647M15.00 0 1996 ElsevierScience Mand Ltd. All rights reserved PU: SO303-2647(96)01612-7