RNAcmap

RNAcmap predicts RNA contact maps by applying evolutionary coupling analysis to derive structural restraints that improve RNA secondary and tertiary structure prediction.

Key Features:

• Fully Automatic Methodology: Automates evolutionary coupling analysis for input RNA sequences without requiring manually curated multiple sequence alignments.
• Broad Applicability: Applies to millions of non-coding RNA sequences rather than being limited to curated families such as Rfam (3,016 families).
• Homology Search via Covariance Models: Constructs covariance models with Infernal using secondary-structure predictions from RNAfold and SPOT-RNA to perform homology searches.
• Secondary-Structure-Driven Performance: Predictive performance is more dependent on the accuracy of the secondary structure predictor than on the specific evolutionary coupling tool used.
• Meta Predictor Capability: Combines outputs from SPOT-RNA and RNAfold to select the most effective homologous sequences and improve contact prediction.

Scientific Applications:

• Enhanced Structure Prediction: Provides base-pairing and contact restraints that aid accurate RNA secondary and tertiary structure modeling.
• Comparative Method Evaluation: Enables assessment of how secondary structure predictors (notably SPOT-RNA) affect evolutionary coupling–based contact predictions.
• Large-Scale Evolutionary Analysis: Facilitates evolutionary coupling analysis across extensive non-coding RNA sequence datasets to identify conserved contacts.

Methodology:

Performs homology searches using covariance models built with Infernal, where covariance models are constructed from secondary-structure predictions produced by RNAfold and SPOT-RNA, followed by evolutionary coupling analysis to infer contacts; the process is fully automated for input RNA sequences.