Helicobacter pylori drug resistance diagnosis system combining machine learning and MALDI-TOF MS

The invention discloses a helicobacter pylori drug resistance diagnosis system combining machine learning and MALDI-TOF MS (matrix-assisted laser desorption ionization-time of flight mass spectrometry). According to the system, on the basis of an MALDI-TOF MS proteomics database of HP clinical strains in China, machine learning prediction models for drug resistance of the strains to the two antibiotics of clarithromycin and levofloxacin are constructed through a LightGBM algorithm, the AUROC of the models to the clarithromycin and the AUROC of the models to the levofloxacin are 0.82 and 0.86 respectively, and the accruities of the models to the clarithromycin and the levofloxacin are 0.75 and 0.72 respectively. And the HPrespred calls models for a plurality of pieces of protein fingerprint data of the HP, and intelligently counts and outputs a drug resistance prediction result. Compared with the prior art, the HPrespred system has the advantages of being short in time consumption, low in cost, small in strain