We utilize the flexible tile model presented in [13] to design self-assembling DNA structures from a graph theory perspective. These tiles represent branched junction molecules whose arms are double strands of DNA.
We consider \( 2 \times n \) triangular lattice graphs \( G_n \), where \( n \) represents the number of triangles. Given a target graph \( G_n \), we determine the minimum number of tile and bond-edge types needed in order to create \( G_n \) as a complete self-assembled complex in three different scenarios. Each scenario corresponds to a distinct level of laboratory constraint.
In the first scenario, graphs of a smaller size than \( G_n \) are allowed. In the second scenario, non-isomorphic graphs of the same size as \( G_n \) are allowed, but not graphs of smaller size. In the third scenario, only graphs isomorphic or larger in size to the target graph are allowed.
We provide optimal tile sets for all \( 2 \times n \) triangular lattice graphs \( G_n \) in Scenario 1 and Scenario 3. We also include some small examples in Scenario 2.