Standardization and Denoising Algorithms for Mass Spectra to Classify Whole-organism Bacterial Specimens

• Activity Number: 427
• Type: Contributed
• Date/Time: Thursday, August 12, 2004 : 10:30 AM to 12:20 PM
• Sponsor: General Methodology
• Author(s): Somnath Datta*+ and Glen Satten and Hercules Moura and Adrian Woolfitt and John Barr
• Companies: University of Georgia and Centers for Disease Control and Prevention and Centers for Disease Control and Prevention and Centers for Disease Control and Prevention and Centers for Disease Control and Prevention
• Keywords: mass spectrometry ; normalization ; pre=processing ; Random Forest ; classification

Abstract

Application of mass spectrometry in proteomics is a breakthrough in high-throughput analyses. Early applications have focused on protein expression profiles to differentiate amongst various types of tissue samples (e.g., normal vs. tumor). We use mass spectra to differentiate between whole-organism samples of bacteria. The raw spectra are similar to spectra of tissue samples, raising some of the same statistical issues (e.g., nonuniform baselines and higher noise associated with higher baseline), but are substantially noisier. As a result, new pre-processing procedures are required before these spectra can be used for statistical classification. We introduce novel pre-processing steps that can be used with any mass spectra. These include a standardization step and a denoising step. The noise level for each spectrum is determined using only data from that spectrum. After applying these preprocessing steps, we used the Random Forest program to classify 120 mass spectra into four bacterial types. The method resulted in extremely low prediction errors in the training samples and zero prediction error in a test dataset we created from the whole dataset.