Symmetry-adapted digital modeling I. Axial symmetric proteins

Considered are axial symmetric proteins exemplified by the octameric mitochondrial creatine kinase, the Pyr RNA-binding attenuation protein, the D-aminopeptidase and the cyclophilin A-cyclosporin complex, with tetragonal (422), trigonal (32), pentagonal (52) and pentagonal (52) point-group symmetry, respectively. One starts from the protein enclosing form, which is characterized by vertices at points of a lattice (the form lattice) whose dimension depends on the point group. This allows the indexing of Cα's at extreme radial positions. The indexing is extended to additional residues on the basis of a finer lattice, the digital modeling lattice Λ, which includes the form lattice as a sublattice. This leads to a coarse-grained description of the protein. In the crystallographic point-group case, the planar indices are obtained from a projection of atomic positions along the rotation axis, taken as the z axis. The planar indices of a Cα are then those of the nearest projected lattice point. In the non-crystallographic case, low indices are an additional requirement. The coarse-grained bead follows from the condition imposed on the residues selected to have a z coordinate within a band of value δ above and below the height of lattice points. The choice of δ permits a variation of the coarse-grained bead model. For example, the value δ = 0.5 leads to a fine-grained indexing of the full set of residues, whereas with δ = 0.25 one gets a coarse-grained model which includes only about half of these residues. Within this procedure, the indexing of the Cα only depends on the choice of the digital modeling lattice and not on the value of δ. The characteristics which distinguish the present approach from other coarse-grained models of proteins on lattices are summarized at the end.