Software DrDMASS+ has been developed to effectively analyze mass spectral data based on multivariate analysis. A flow diagram of Data processing consists of four stages, (i) Peak Correction, (ii) Multivariate Data Processing, (iii) Unsupervised Learning, and (iv) Supervised Learning. In Peak Correction process, we correct experimental m/z values based on the relation between experimental and desired values of internal mass calibrants (IMCs). A multivariate data is consisting of a data set of multiple samples. In Multivariate Data Preprocessing, we can assess reproducibility of samples with iterative measurement, and select useful peaks for separating groups of samples and so on. In Unsupervised Learning, we can visualize the multivariate data by using multivariate analysis method such as principal component analysis (PCA) and Batch-learning self-organizing map (BL-SOM). In Supervised Learning, we can get the regression equation by using Partial Least Squares Regression (PLS).
Copyright (C) 2007 NAIST Comparative Genomics Laboratory All Rights Reserved.