DNA contains information beyond that in its arrangement of A, C, G, T bases. The structure of the DNA (how bases are twisted relative to each other, how deep the grooves between the DNA backbones are, etc.) can have a large impact on vital genomic processes, such as replication, transcription, and nucleosome packaging. It is possible to predict DNA structure from sequence. Databases exist that assign numeric values for various DNA structural properties to dinucleotides. In this paper we leverage that information to explore the space of sequences similar to six reference sequences in terms of 25 structural properties. This is a necessary prerequisite to future work that will develop a computational process that characterizes DNA sequences with specific biological functions based on similar structural properties in a way analogous to the way sequence logos characterize groups based on sequence. A consequence of this study is insight into the relationships between the various structural properties. Additionally, we discover serendipitous fitness functions that find sequences with properties they were not designed to discover.
