Fig. 1
A maximum-likelihood tree of the 2HADHs from 111 representative organisms. The tree was computed with FastTree 2.1.7 [106] based on a high-quality, structure-based multiple sequence alignment and visualized with Archaeopteryx [108]. The separated subfamilies were defined based on high support values of the corresponding bifurcations and consistency between trees computed using different approaches. Proteins studied biochemically are marked with circles, which denote their substrates (large, most efficient in terms of kcat/KM; small, additional). SERA, 3-phosphoglycerate dehydrogenases; FDH, formate dehydrogenases; CTBP, C-terminal binding proteins; PDXB, 4-phosphoerythronate dehydrogenase; LDHD, D-lactate dehydrogenases; GHRA, glyoxylate/hydroxypyruvate reductases A; GHRB, glyoxylate/hydroxypyruvate reductases B; GHRC, glyoxylate/hydroxypyruvate reductases C; DDH, broad-substrate-specificity dehydrogenases; and X1-X13, subfamilies not studied biochemically. Nodes with local support values greater than 0.8 are denoted by grey squares. The tree in Newick format with branch support values can be found in Additional file 3: Data file S1