Towards parsimonious generative modeling of RNA families
Résumé
Generative probabilistic models emerge as a new paradigm in data-driv en, e v olution-inf ormed design of biomolecular sequences. This paper introduces a no v el approach, called Edge A ctiv ation Direct Coupling Analy sis (eaDCA), tailored to the characteristics of RNA sequences, with a strong emphasis on simplicity , efficiency , and interpretability . eaDCA explicitly constructs sparse coevolutionary models for RNA families, achie ving perf ormance le v els comparable to more comple x methods while utilizing a significantly lo w er number of parameters. Our approach demonstrates efficiency in generating artificial RNA sequences that closely resemble their natural counterparts in both statistical analyses and SHAPE-MaP experiments, and in predicting the effect of mut ations. Not ably, eaDCA provides a unique feature: estimating the number of potential functional sequences within a given RNA family. For example, in the case of cyclic di-AMP riboswitches (RF00379), our analysis suggests the e xistence of appro ximately 10 39 functional nucleotide sequences. While huge compared to the known < 40 0 0 natural sequences, this number represents only a tiny fraction of the vast pool of nearly 10 82 possible nucleotide sequences of the same length (136 nucleotides). These results underscore the promise of sparse and interpretable generative models, such as eaDCA, in enhancing our understanding of the e xpansiv e RNA sequence space.
Domaines
| Origine | Fichiers éditeurs autorisés sur une archive ouverte |
|---|---|
| Licence |
![]()
Cite 10.5281/zenodo.10688226 Ouvrage FrancescoCalvanese. (2024). FrancescoCalvanese/FCSeqTools.jl: Towards parsimonious generative modeling of RNA families (Version v1.0.0). Zenodo. https://doi.org/10.5281/ZENODO.10688226