SARA

SARA aligns RNA structures pairwise to enable structure-based assignment of RNA function.

Key Features:

• Alignment algorithm: Aligns two RNA structures using a unit-vector root-mean-square approach for structural superposition.
• Multi-level information: Considers sequence information together with secondary- and tertiary-level structural information during alignment.
• Statistical scoring: Computes three distinct P-values that estimate the statistical significance of sequence identity, secondary-structure identity, and tertiary-structure identity scores.
• Benchmark validation: Validated on a benchmark set of 419 RNA structures with known SCOR structural classes, reporting that −log(mean P-value) ≥ 2.5 yields 81.4% correct original SCOR-class assignments and 95.3% correct similar SCOR-class assignments.
• Relevant datasets: Applicable to analysis of non-coding RNA transcripts and RNA structures deposited in the Protein Data Bank.

Scientific Applications:

• Structure-based function assignment: Assigns putative functions to RNA molecules by structural similarity to annotated RNAs.
• Functional classification: Classifies RNA structures into SCOR structural classes based on pairwise structural comparisons.
• Comparative structural analysis: Enables comparative analysis of secondary and tertiary structural features among RNA molecules, including non-coding RNAs.

Methodology:

Performs pairwise alignment of two RNA structures using a unit-vector root-mean-square approach, computes sequence, secondary-structure and tertiary-structure identity scores, derives three corresponding P-values and a mean P-value whose negative logarithm is used to assess assignment significance; benchmarking used 419 SCOR-classified RNA structures.