Figure 4: Comparison of the dendrogram of whole cell protein profiles and 16S rDNA-based phylogenetic trees of rhizobia. a) The dendrogram was generated via PAUP 4.0b package. Briefly, totally 15-18 replicates of protein peak lists (3 biological replicates and 6 technical replicates from each sample) were combined together if the same protein peaks repeated greater than 95% of chances in all 15-18 replicates that were counted. Each sample was, thus, assigned to “mainspectra peak lists”. Subsequently, we coded all the sample’s peak lists into a binary classifier system (present 1, absent 0). The generated main data matrix was used to perform phylogenetic analysis via PAUP 4.0b. The distance matrix tree was calculated with UPGMA method, with 1000 bootstrap value. b), the phylogenetic tree was generated using full length of 16S rDNA sequences, the tree was also calculated with UPGMA method by using MEGA 5.0 package. An accession number (GenBank No.) was listed. The UPGMA tree was tested with the bootstrap method (bootstrap replicates 1000). The percentage of replicates in which the associated taxa clustered together in the bootstrap test is shown next to the branches. The value less than 50 indicated the unstable clade and the branches might collapse. The substitution model used nucleotide type and maximum composition likelihood method. The branches unit is the number of base substitution per site.