Methods and applications of RNA contact prediction*

Abstract

The RNA tertiary structure is essential to understanding the function and biological processes. Unfortunately, it is still challenging to determine the large RNA structure from direct experimentation or computational modeling. One promising approach is first to predict the tertiary contacts and then use the contacts as constraints to model the structure. The RNA structure modeling depends on the contact prediction accuracy. Although many contact prediction methods have been developed in the protein field, there are only several contact prediction methods in the RNA field at present. Here, we first review the theoretical basis and test the performances of recent RNA contact prediction methods for tertiary structure and complex modeling problems. Then, we summarize the advantages and limitations of these RNA contact prediction methods. We suggest some future directions for this rapidly expanding field in the last.

Salt-Dependent RNA Pseudoknot Stability: Effect of Spatial Confinement

“…Although we made reliable predictions and comprehensive analyses on salt-dependent structure stability for the MMTV pseudoknot under different spatial confinements, the effect of spatial confinement on structure stability for extensive RNA pseudoknots with various sequences/lengths and functional RNAs with complex structures (e.g., tRNA and riboswitches) in ion solutions should be studied by further developing the present CG model (Zhang et al, 2011;Kilburn et al, 2016;Yamagami et al, 2018), and further development on the present model may need to take into account the tertiary contacts beyond the canonical and wobble base pairs already involved in the model (Weinreb et al, 2016;Jian et al, 2019;Wang and Zhao, 2020). Furthermore, there are some other physical and chemical factors that could affect how crowding modulates RNA stability, which is beyond the approximation of the spatial confinement considered here (Zhou et al, 2008;Nakano et al, 2009;Paudel and Rueda, 2014;Strulson et al, 2014;Leamy et al, 2016Leamy et al, , 2017Feig et al, 2017).…”

Resident- and Institutional-Level Factors, Frailty, and Nursing Homes Residents

“…The advantages of multilevel models are provided in Supplementary Digital Content (Text 1, http://links.lww.com/NRES/A411). In building our final multilevel models, we followed the recommended procedures to answer our research questions (Wang et al, 2008), which are described further in the Results section. Model fit was evaluated using the Akaike information criterion and the Bayesian information criterion.…”

1D2DSimScore: A novel method for comparing contacts in biomacromolecules and their complexes

“…However, for nucleic acids (in particular RNA) and protein–nucleic acid complexes, the generation of 3D models is not yet a routine procedure. On the other hand, numerous tools have been developed for predicting contacts between proteins and nucleic acids 17,18 . Besides, the most common way of describing RNA structure is in the form of 2D contacts, which may range from only canonical base pairs as in RNA secondary structure, or may also involve non‐canonical pairs and stacking interactions.…”

“…On the other hand, numerous tools have been developed for predicting contacts between proteins and nucleic acids. 17 , 18 Besides, the most common way of describing RNA structure is in the form of 2D contacts, which may range from only canonical base pairs as in RNA secondary structure, or may also involve non‐canonical pairs and stacking interactions. There exist numerous tools for extracting contact information from proteins and nucleic acid structures, for example, PROTMAP2D, 19 RNAMAP2D, 20 DSSR, 21 and ClaRNA, 22 which facilitate the comparison of structural information for macromolecules with known 3D structures, and the ones for which only the contact data are available.…”