The mass spectrometry spectra are recognized as a screening tool for detecting discriminatory protein patterns. However, the mass spectra represent high dimensional data that have a large number of local maxima (a.k.a. peaks) which have to be analyzed; to tackle this problem we have developeda new three-step strategy. After preprocessing for classification of mass spectra, we use analgorithm clonal selection for synthesis collective binary classifiers in the form of wavelet-neural networks. The results obtained by the analysis of a data set of tumor/healthy samples allowed us to correctly classify more than 99% of samples.