This work discloses a unique, comprehensive proteomic dataset of Acinetobacter baumannii strains, both resistant and non-resistant to polymyxin B, isolated in Brazil generated using Orbitrap Fusion Lumos. From nearly 4 million tandem mass spectra, the software DiagnoMass produced 240,685 quality-filtered mass spectral clusters, of which PatternLab for proteomics identified 44,553 peptides mapping to 3479 proteins. Crucially, DiagnoMass shortlisted 3550 and 1408 unique mass spectral clusters for the resistant and non-resistant strains, respectively, with only about a third with sequences (and PTMs) identified by PatternLab. Further open-search attempts via FragPipe yielded an additional ~20% identifications, suggesting the remaining unidentified spectra likely arise from complex combinations of post-translational modifications and amino-acid substitutions. This highlights the untapped potential of the dataset for future discoveries, particularly given the importance of PTMs, which remain elusive to nucleotide sequencing approaches but are crucial for understanding biological mechanisms. Our innovative approach extends beyond the identifications that are typically subjected to the bias of a search engine; we discern which spectral clusters are differential and subject them to increased scrutiny, akin to spectral library matching by comparing captured spectra to themselves. Our analysis reveals adaptations in the resistant strain, including enhanced detoxification, altered protein synthesis, and metabolic adjustments. Significance: We present comprehensive proteomic profiles of non-resistant and resistant Acinetobacter baumannii from Brazilian Hospitals strains, and highlight the presence of discriminative and yet unidentified mass spectral clusters. Our work emphasizes the importance of exploring this overlooked data, as it could hold the key to understanding the complex dynamics of antibiotic resistance. This approach not only informs antimicrobial stewardship efforts but also paves the way for the development of innovative diagnostic tools. Thus, our findings have profound implications for the field, as far as methods for providing a new perspective on diagnosing antibiotic resistance as well as classifying proteomes in general.