Structural Bioinformatics comprises the use of large datasets of structural three-dimensional data of primarily proteins. Study of these datasets allows the ab initio calculation of properties and vice versa, similar technologies can be used to predict structures of biologically active molecules and of their interactions. Thus a true structure-activity relationship can be constructed for proteins and their interactions with ligands, this includes predictions on the non-static behavior of proteins, which are known to be represented by more than one single 3D structure.

Commonly used techniques are structure alignments, phylogenetic analysis, homology modeling, structure analysis, molecular dynamics simulation, binding free energy calculations, protein docking, virtual screening.