Skip to main content

Table 9 Sample classification after LDA to discriminate between CSF samples from patients with Relapse–remitting multiple sclerosis (RRMS) vs. other diseases (inflammatory and non-inflammatory) of the CNS (INF and N-INF) based on the identified protein spots

From: Posttranslational modifications of proteins are key features in the identification of CSF biomarkers of multiple sclerosis

  Classification resultsb,c
Disease Predicted Group Membership Total
RRMS INF and N-INF
Original
 Count RRMS 57 12 69
INF and N-INF 11 58 69
 % RRMS 82.6 17.4 100
INF 15.9 84.1 100
Cross-validateda
 Count RRMS 56 13 69
INF and N-INF 11 58 69
 % RRMS 81.2 18.8 100
INF & N-INF 15.9 84.1 100
  1. aCross validation is done only for those cases in the analysis. In cross validation, each sample is classified by the functions derived from all samples other than the one under analysis
  2. b83.3% of original grouped cases correctly classified
  3. c82.6% of cross-validated grouped cases correctly classified