Skip to main content

Small ribonucleic acids in silico

Final Report Summary - S-RNA-S (Small ribonucleic acids in silico)

Ribonucleic acid (RNA) has a crucial role in the post-transcriptional regulation of gene expression. One goal of this project was to bring state-of-the-art techniques used in the molecular dynamics community into the RNA field. In particular, our group had a native expertise in enhanced sampling methods. Another goal was to apply these techniques, in combination with standard molecular dynamics and state-of-the-art rare-event methods, to the study of topics including riboswitch folding and dynamics, RNA-protein interaction and folding of non-coding RNAs.

In these years we have published a number of methodological papers describing new techniques that can be used to either accelerate molecular dynamics simulations or to analyze them. All these techniques have been made available to the scientific community. A major result of our project was the publication of a novel version of the software PLUMED, developed in collaboration with other research groups. It is relevant to mention that many of the techniques developed in our group were already used by other groups. We believe that the techniques we developed represent a major improvement beyond the state of the art and will influence the RNA-simulation community in the next years.

At the same time, we applied these techniques to a number of biologically relevant problems. We performed molecular dynamics simulations of a RNA-helicase complex. Simulation studies on RNA-protein complexes are still very rare, and we believe we gave a proof of concept of what can be done using these techniques. In addition, we applied our methods to the characterization of riboswitch folding, focusing on the effect of ligand binding on the conformational landscape. Our results clarify how ligand and ions influence riboswitch dynamics and, ultimately, allow gene expression to be controlled. Finally, we used enhanced sampling methods, in combination with techniques to combine molecular dynamics simulations with NMR data, to characterize the structural dynamics of a non-coding RNA.

In addition to these works, we had to solve an unforeseen problem. Namely, the quality of the models used in molecular dynamics simulations turned out to be insufficient for the blind theoretical prediction of experimental results. We tackled this problem in practice by combining molecular dynamics simulations with experimental data and by developing a “knowledge based” coarse grained methods. Whereas the quality of models used in molecular dynamics simulations was not clear when we started our project, the work from our and other groups clearly demonstrated how far blind prediction of RNA structure can be pushed, establishing significant benchmarks. We believe these benchmarks will have a significant influence in future research.

Overall, the main objectives of the project were accomplished.