| | TM1 | TM2 | TM3 | TM4 | TM5 |
|---|
TPT-1/DUF914-1
| - | P | L(1)L | - | G(7)D |
DUF914-1/UAA-1
| - | F | - | A(2)Y | Y(13)GV |
UAA-1/NST-1
| - | - | - | - | L(3)GV |
EamA-1/EamA-2
| L(1)K(2)L | - | I(3)G | - | I(1)G(6)G |
EamA-2/TPT-2
| G(4)L(6)AL(2)V(2)K | - | - | SV | G(6)G(7)K |
TPT-2/DUF914-2
| G(4)L(6)A(3)V | - | F | TS | G(6)G(3)Y |
DUF914-2/UAA-2
| G(2)L(8)A(6)E | - | - | I(6)T(6)L | G(6)G(2)L |
UAA-2/NST-2
| G(14)G(3)E | - | - | T(10)S(2)LF(4)T | G(2)IV(5)L |
- Pairwise comparisons are made for HMM consensus sequences of DMT families that are presumed, from Results, to have evolved from each other (pairs are listed in 1st column). The aligning residues in the consensus sequences are obtained from HHsearch. Aligning HHsearch consensus residues (representing ~33% sequence conservation) are counted if they are located in TM segments. In the first TM segment of the second domain, a motif G-X(11)-A is found, and in the 5th TM segment of the second domain, G-X(6)-G is discovered. Some of the TPT-1 sequences are EamA sequences, because TPT is not found in the DMT-1 slot; we assume that DUF914-1 evolved from EamA-1 directly.
