Biological systems are widely regarded as performing computations. It is much less clear, however, what exactly is computed and how biological computation fits within the framework of standard computer science. Here we explore the idea that evolution confines biological computation to subsets of instances that can be solved efficiently with algorithms that are ‘hardcoded’ in the system itself. We use RNA secondary structure prediction as a simple surrogate for developmental programs to demonstrate that the salient features of the genotype–phenotype map remain intact even if ‘simpler’ algorithms are employed that correctly compute the structures only for small subsets of instances, albeit quantitative differences depending on the choice of alternative algorithms can be observed.

Mechanically active RNAs represent an emerging class of biomolecules whose function derives from resisting molecular forces. Among them, exoribonuclease-resistant RNAs (xrRNAs) achieve this by folding into a ring-like topology that physically blocks $5^{\prime } \rightarrow 3^{\prime }$ degradation. However, despite years of structural insight, the rational design of such mechanically functional RNA devices has remained elusive. Here, we describe a mechanics-aware RNA design approach that enables de novo engineering of functional xrRNAs. We first identify structural determinants of force resistance by perturbing pseudoknot architecture in a model xrRNA and quantifying resulting efficiencies in the stalling of exoribonuclease XRN1. We then implement these rules in a design framework that integrates explicit topological constraints with molecular dynamics-guided optimization. The resulting synthetic xrRNAs reproduce the ring-like architecture and stall exoribonuclease XRN1 with wild-type-like efficiency. Our top-performing constructs exhibit minimal sequence similarity to known xrRNAs and evade detection by covariance models, yet remain fully functional in vitro. Together, our results show that mechanical function can be rationally designed independent of evolutionary ancestry, laying the groundwork for the design of RNA elements that modulate decay and fine-tune the mechanical stability of engineered transcripts.

RNA aptamers are often used as ligand-recognition modules in engineered RNA systems, but integration into larger RNA constructs can influence stability and ligand binding. As a result, aptamer sequences may need to be adapted to new environments while preserving essential properties. Here, we examine this sequence editability problem for the theophylline RNA aptamer. Starting from the experimentally determined structure, we introduced targeted mutations in peripheral structural elements while leaving the recognition site unchanged. The native aptamer, mutated variants, a Mg²⁺-depleted system, and a caffeine-bound control were analyzed using three independent 1 $\mathsf{\mu }$s molecular dynamics simulations. Binding energetics were estimated with multiple end-point as well as alchemical free energy approaches. Results were interpreted together with base pair stability, the conformational landscape of the binding pocket, and per-nucleotide energy contributions. This allows us to predict whether an edit is tolerated or disruptive. Some mutations retained structural and energetic profiles comparable to the native aptamer, whereas others reduced ligand affinity by propagating structural distortions into the binding pocket. These results show that sequence changes outside the binding site can modulate ligand binding indirectly, and that the ligand interaction network is useful for evaluating edited aptamers. The introduced workflow provides a novel combination of established computational strategies for efficient in silico screening of aptamer variants before experimental testing and can be integrated into the design of larger RNA structures. This works particularly well when an experimental structure is available and the tested mutations are small enough not to disrupt the folding pathway.

In recent years, several methods for detecting RNA-RNA interactions have become available that use a combination of crosslinking, ligation, and sequencing of the resulting chimeric reads. In principle, such data also convey information on intramolecular helices. They are, however, not accurate enough to identify base pairs directly. Instead, only regions of direct contacts can be inferred. Here, we show that such data can be incorporated as pseudo-energies into RNA secondary structure prediction algorithms by assigning a bonus term to all potential pairs between crosslinked intervals. Using simulated data, we show that given sufficient coverage, such data can push the accuracy of the predicted structure to a base pair-wise MCC of above 90%. Moreover, we observe that the beneficial effect of such interval-wise pseudo-energies is quite robust w.r.t. the length of the interval and the value of the bonus term, but depends strongly on the fraction of the sequence that is covered by significant interaction data.