Exploratory study on application of MALDI-TOF-MS to detect serum and urine peptides related to small cell lung carcinoma

ClinProToolsTM (CPT) software was used for the statistical analysis of all mass spectral data obtained. Before complete analysis using CPT, the original mass spectra of the samples were processed, including baseline correction, alignment, and normalization. Next, the statistical analysis of the serum peptide spectra of the 54 patients with SCLC and 54 healthy individuals composing the training group was conducted using the statistical algorithms included in CPT. The peptides differentially expressed between the two groups were obtained.Three algorithms (genetic algorithm [GA], supervised neural networks [SNN], and quick classifier [QC]) were used to establish the prediction models. Finally, the 18 patients with SCLC and 18 healthy individuals of the testing group were used for blinded validation of the classification model, to validate the stability and reliability of the model. Similarly, the statistical analysis of urine peptides revealed the urine peptides differentially expressed between the 54 patients with SCLC and 54 healthy individuals in the training group, and a urine classification model for SCLC diagnosis was constructed from a number of peptides selected by CPT. Finally, the 18 patients with SCLC and 18 healthy individuals in the testing set were used for blinded validation of the classification model, to validate the stability and reliability of the model