A mathematical model for classical chromosome identification using the logical combinatory approach

The objective of this work was to develop a mathematical model to describe and evaluate the chromosomes which arises in the cytogenetic classification problem, using the logical combinatory approach of pattern recognition theory. Based on the classic chromosome identification approach, seven variables with their domain and comparison criteria were defined, as well as an analogy function that allows to evaluate the similarity between two chromosome descriptions. Based on this model, a computational tool was developed. This tool identifies the chromosomes by calculating the similarity between them. The tool is able to perform three kinds of identification, a) Exact identification, when the chromosome matches just with only one other chromosome, b) Multiple identification, when the chromosome matches with more than one chromosome; c) Approximated identification, when the chromosome identification does not match with any chromosome and it is, therefore, approximated by the similarity factor. Considering these three identification strategies, the computational tool can identify all the twenty four known chromosome couples.