Browsing by Subject "Heuristic algorithms"

DNA origami has been a paradigm-shifting technique for synthesizing and manipulating matter with nanoscale precision. The simple design principle of using numerous short (<100 nts) oligonucleotides to "fold" a long (>1000 nts) DNA strand achieved both simplicity in design and greatly increased yields in comparison to previous motifs for DNA nanostructure design. Various approaches have been explored that have resulted in DNA nanostructures rapidly growing in mass and complexity while also becoming more accessible for a wide scientific community, such as developing computer-aided design graphical user interfaces, establishing design principles for classes of structures with algorithmic regularity, and refining synthesis strategies and the respective design criteria to exploit them.

These directions are all fundamentally a straight extension of the DNA origami technique and pursuits towards large, functional DNA origami have been amply rewarded. Yet due to the nature of how a primary driving factor of scaling designs upwards has been the exploitation of repeatable motifs, several assumptions underlie conventional strategies for the DNA origami design of complex shapes. This thesis formally classifies a geometry of curved DNA origami nanostructures and discusses how such structures do not align with existing assumptions for DNA nanostructure design. While it is class of structures that has high biotechnological relevance, the tedium of design challenges arising from this departure have limited accessibility and enthusiasm for utilizing them. To achieve greater functional relevance, DNA origami must undoubtedly retread on the establishment of strategies for scaling up mass and shape complexity in DNA nanostructures; this time beyond regular, repeating subunits, and towards supramolecular assemblies with distinct, bespoke geometric features. As such, this thesis entreats an approach towards formalizing local and global properties in DNA origami design that can be quantified and characterized for their effects on DNA nanostructure yield and stability. Thus, a generalized strategy for DNA origami design can be born.

This thesis first consolidates and proposes a hierarchy of properties active in DNA origami design. It then suggests and evaluates two heuristic optimization algorithms to attempt a multi-variable optimization of those properties to achieve rapid generation of oligonucleotide sequences to generate desired DNA origami shapes. This thesis then discusses the existing challenges and potential applications of curved DNA origami nanostructures. Lastly, the application of the aforementioned optimization algorithms are applied to generate examples in this class of nanostructures, and the results are hither reported and discussed.