In the last years scientists all around the world have been obtaining the complete genomic sequences from different organisms. The human genome , for instance, contains about three thousand million ( 3x10exp9) bp (base pairs). The great ammount of information we already have will force the researchers, in these next years to come, to make a considerable effort in the comprehension of the hidden clues lying behind this “naked” information. Actually, the path that will lead from information to real knowledge needs to be walked.

GENE PREDICTION in DNA sequencies is an essential goal in the understanding of the annotated genome of any organism. Only the 2% of the human genome is proteďn-coding. Thus, inside this little percentage we must look for the genes that from which a certain organism develops.

There are programmes that enables us to predict the exons contained inside a given DNA sequence ( GeneID ). This programmes, which are useful tools, perform a degenerated prediction of exons. This means that they predict a larger number of exons than the total number of them really existing in the sequence. The explanation to this fact is very easy: the models that these programmes use work only with the DNA sequence reducing the genetic information to the symbols A, T, G, C. They are valid theoretical models in the attempt of reaching genetic information complexity although they surely miss some outstanding issues like 3-D conformation or trans elements associated to dsDNA. Including this issues in the modelitzation of genetic information would extremely complicate its analysis (and this is main the reason why they’re not included). However, the modelitzation based in the four different kind of nitrogenated bases (A, T, G, C) is very powerfull. We only need to be criticals enough according to the information we know we ignore in our models of nature.

We developed a programme named CERCAGEN , which is complementary to the programmes we mentioned before. Starting with a collection of predicted exons, CERCAGEN will calculate, according to the associated score of each single exon , which is the gene that will emerge with the higher probability from the total of exons predicted.