Among the many applications of mass spectrometry, biomarker pattern discovery from protein mass spectra has aroused huge interest in the recent years. While research efforts have raised hopes of early and less invasive diagnosis, they have also brought to light the many issues to be tackled before mass-spectra-based proteomic patterns become routine clinical tools. Undoubtedly, biomarker selection among the high dimensional input data is the most critical part of each pattern recognition algorithm applied to this area. In this paper we pursued a new feature selection strategy that explores all data points as initial features rather than just peaks. Using the derived features in conjunction with only two intuitive fuzzy rules, we achieved a considerable accuracy over a couple of well-known ovarian cancer datasets