Rationally designed split Lettuce aptamer based on large scale mutational analysis

Split aptamer biosensors offer exceptionally low background by assembling only in the presence of a target analyte; however, their performance is frequently limited by the lack of robust design rules for selecting effective split sites. Existing approaches largely rely on heuristic, structure-based assumptions that are poorly validated and often yield suboptimal signal. Herein, we introduce a systematic, data-driven strategy for identifying high-performance split sites within fluorogenic DNA aptamers. Using our massively-parallel aptamer performance analyzer (MAPA) platform, we performed comprehensive single- and double-mutant analysis of the DFAME-binding region of the fluorogenic DNA aptamer Lettuce, informed by its three-dimensional structure. Dimensionality reduction and clustering of the resulting sequence-function landscape revealed mutation-tolerant elements within the binding domain that are suitable for splitting while preserving fluorophore activation. Sensors constructed using these non-intuitive split sites, which are unconventional by standard design principles, exhibited a nearly four-fold improvement in fluorescence signal-to-background ratio for SARS-CoV-2 RNA detection compared to a prior split-Lettuce design. The same split architecture also enabled robust detection of high-pathogenicity H5Nx avian influenza RNA. These results demonstrate that large-scale, data-driven interrogation of aptamer sequence-function relationships can identify non-intuitive split sites and provide a proof-of-concept framework for developing measurement-based design principles for split-aptamer biosensors.