Physica A 348 (2005) 349–370 A recursive network approach can identify constitutive regulatory circuits in gene expression data Monica Francesca Blasi a , Ida Casorelli a , Alfredo Colosimo b , Francesco Simone Blasi c , Margherita Bignami a , Alessandro Giuliani a,Ã a Department of Environment and Primary Prevention, Istituto Superiore di Sanita`, Viale Regina Elena 299, Rome 00161, Italy b Department of Human Physiology and Pharmacology, Universita` di Roma ‘‘La Sapienza’’, Rome, Italy c Department of Mathematics, Universita’ di Roma ‘‘Tor Vergata’’, Rome, Italy Received 14 July 2004 Available online 10 November 2004 Abstract The activity of the cell is often coordinated by the organisation of proteins into regulatory circuits that share a common function. Genome-wide expression profiles might contain important information on these circuits. Current approaches for the analysis of gene expression data include clustering the individual expression measurements and relating them to biological functions as well as modelling and simulation of gene regulation processes by additional computer tools. The identification of the regulative programmes from microarray experiments is limited, however, by the intrinsic difficulty of linear methods to detect low- variance signals and by the sensitivity of the different approaches. Here we face the problem of recognising invariant patterns of correlations among gene expression reminiscent of regulation circuits. We demonstrate that a recursive neural network approach can identify genetic regulation circuits from expression data for ribosomal and genome stability genes. The proposed method, by greatly enhancing the sensitivity of microarray studies, allows the identification of important aspects of genetic regulation networks and might be useful for the discrimination of the different players involved in regulation circuits. Our results suggest that ARTICLE IN PRESS www.elsevier.com/locate/physa 0378-4371/$-see front matter r 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.physa.2004.09.005 Ã Corresponding author. Tel.: +390649902579; fax: +390649902255. E-mail address: alessandro.giuliani@iss.it (A. Giuliani).