The energy-spectrum of bicompatible sequences

Fenix W. Huang, Christopher L. Barrett, Christian M. Reidys
2021 Algorithms for Molecular Biology  
Background Genotype-phenotype maps provide a meaningful filtration of sequence space and RNA secondary structures are particular such phenotypes. Compatible sequences, which satisfy the base-pairing constraints of a given RNA structure, play an important role in the context of neutral evolution. Sequences that are simultaneously compatible with two given structures (bicompatible sequences), are beacons in phenotypic transitions, induced by erroneously replicating populations of RNA sequences.
more » ... A riboswitches, which are capable of expressing two distinct secondary structures without changing the underlying sequence, are one example of bicompatible sequences in living organisms. Results We present a full loop energy model Boltzmann sampler of bicompatible sequences for pairs of structures. The sequence sampler employs a dynamic programming routine whose time complexity is polynomial when assuming the maximum number of exposed vertices, $$\kappa $$ κ , is a constant. The parameter $$\kappa $$ κ depends on the two structures and can be very large. We introduce a novel topological framework encapsulating the relations between loops that sheds light on the understanding of $$\kappa $$ κ . Based on this framework, we give an algorithm to sample sequences with minimum $$\kappa $$ κ on a particular topologically classified case as well as giving hints to the solution in the other cases. As a result, we utilize our sequence sampler to study some established riboswitches. Conclusion Our analysis of riboswitch sequences shows that a pair of structures needs to satisfy key properties in order to facilitate phenotypic transitions and that pairs of random structures are unlikely to do so. Our analysis observes a distinct signature of riboswitch sequences, suggesting a new criterion for identifying native sequences and sequences subjected to evolutionary pressure. Our free software is available at: https://github.com/FenixHuang667/Bifold.
doi:10.1186/s13015-021-00187-4 pmid:34074304 pmcid:PMC8167974 fatcat:3a3kf5xjkbagpovlsvfse57rum