Discrimination of single amino acid mutations of the p53 protein by means of deterministic singularities of recurrence quantification analysis

p53 is mutated in roughly 50% of all human tumors, predominantly in the DNA- binding domain codons. Structural, biochemical, and functional studies have reported that the differ- ent p53 mutants possess a broad range of behaviors that include the elimination of the tumor-suppres- sion function of wild-type protein, the acquisition of dominant-negative function over the wild-type form, and the establishment of gain-of-function activities. The contribution of each of these types of mutations to tumor progression, grade of malignancy, and response to anticancer treatments has been so far analyzed only for a few "hot-spots." In an attempt to identify new approaches to systematically character- ize the complete spectrum of p53 mutations, we applied recurrence quantification analysis (RQA), a non-linear signal analysis technique, to p53 primary structure. Moving from the study of the p53 hydro- phobicity pattern, which revealed important simi- larities with the singular deterministic structuring of prions, we could statistically discriminate, on a pure amino acid sequence basis, between experimen- tally characterized DNA-contact defective and con- formational p53 mutants with a very high percent- age of success. This result indicates that RQA is a mathematical tool particularly advantageous for the development of a database of p53 mutations that integrates epidemiological data with structural and functional categorizations. Proteins 2004;55:743-755.