Positional Anchors and Cooperative Retained-Position Sets Define Non-Equivalent RNA Sequence Landscapes

Different RNA sequences do not have equal access to nearby structural sequence space, but how strongly local sequence neighborhoods depend on starting-sequence identity and individual nucleotide positions remains unclear. Here, we compared the fully mutated neighborhoods of three isolated 20-nucleotide RNA seeds inspired by the P-site peptidyl transferase center, shifted P-loop, and A-site A-loop regions of the ribosomal core using ViennaRNA minimum-free-energy secondary-structure predictions. The seeds were treated as conceptual starting points for sequence-space exploration rather than exact residue-numbered extracts from a single modern rRNA. The three seeds generated reproducibly different predicted foldability landscapes, with the shifted P-loop-inspired seed consistently producing the highest foldability across independent pools and increasingly stringent MFE thresholds. Single-position retention revealed strong positional effects: retained guanine generally increased predicted foldability, but one retained-G position in each seed significantly outperformed other retained-G positions within the same sequence. Disabling G·U wobble pairing reduced the magnitude of these effects but preserved the retained-base hierarchy, the identity of all three positional anchors, and the overall seed ranking. Combinations of retained positions produced super-additive increases in predicted foldability, although structural-proximity analysis showed that these positions did not consistently form a single compact predicted structural core. Iterative computational selection for lower MFE progressively enriched all three neighborhoods for predicted structure, while six non-conserved biological control 20-mers generated even richer predicted foldability landscapes than the ribosome-inspired seeds. Together, these results show that local RNA sequence space is strongly shaped by starting-sequence identity, nucleotide position, and higher-order retained-position effects and cannot be explained by nucleotide composition or starting-seed structure alone. These conclusions are restricted to MFE-based secondary-structure predictions and do not establish tertiary folding, biochemical function, or experimentally populated structures.