Proteins selected according to the high reactivity of IgG antibodies in the sera of ALS patients.

1Obtained from Life Technologies.2Spot intensity median concerning control samples.3Spot intensity median concerning ALS samples.4Feature frequency within 100 feature selection subruns.5Position of the respective protein feature in the ranked feature list obtained from PAM analysis.6P-value obtained from an independently performed M score computation.7False discovery rate (method: Benjamini-Hochberg) for the M score p-values.8Feature position in the list of all ProtoArray features sorted by means of M score.9P-value obtained from an independently performed t test.10False discovery rate (method: Benjamini-Hochberg) for the t test p-values.Overview of the 20 final selected proteins. The columns “Database ID” and “Description” show the corresponding specifications obtained from Life Technologies. The table is sorted by the means of feature frequency within 100 feature selection subruns (“Frequency”) used to select these reactive antibodies (cut-off: at least 10%). The columns “NDC” and “ALS” show the corresponding group-specific spot intensity medians, and the column ”PAM re-verification” gives the position of the respective protein feature in the ranked feature list obtained from PAM analysis. Additionally, the M scores (“M Score (p value)”) and t-test p values (“t test p value”) obtained from an independently performed 2-group comparison (i.e., not used for feature selection) using all samples (20 ALS vs. 20 NDC), the corresponding adjusted p values (“M Score FDR” and “t-test FDR”, method: Benjamini-Hochberg), and the corresponding positions of the features in the list of all ProtoArray features sorted by means of M score (“M Score Position”) are given for each feature to facilitate understanding of these p values.