CentroidHomfold-LAST

CentroidHomfold-LAST integrates homologous sequence information retrieved by the LAST aligner with the CentroidHomfold algorithm to improve RNA secondary structure prediction accuracy using evolutionary relationships among sequences.

Key Features:

• Homologous sequence integration: Incorporates homologous sequences and evolutionary information into RNA secondary structure prediction to account for conservation across related sequences.
• Automated homolog retrieval: Automatically identifies and retrieves homologous sequences from large databases using the LAST local aligner.
• CentroidHomfold methodology: Applies the CentroidHomfold algorithm to refine target RNA secondary structure predictions based on homologous sequence data.
• Comparative performance: Shows improved predictive accuracy relative to CentroidFold and RNAfold as demonstrated by computational experiments on commonly used datasets.

Scientific Applications:

• Functional annotation: Provides more reliable secondary structure models to support functional annotation of RNAs.
• Comparative genomics: Enables comparative analyses of RNA structures across species by leveraging homologous sequence information.
• Evolutionary studies: Assists evolutionary investigations by revealing conserved structural features using homologs.
• Structure–function and interaction inference: Aids inference of RNA–RNA and RNA–protein interactions and structure–function relationships through improved secondary structure predictions.

Methodology:

Homologous sequences are identified and retrieved from large databases using LAST, these homologs are input to the CentroidHomfold algorithm to refine target RNA secondary structure predictions, and performance is evaluated via computational experiments comparing results to CentroidFold and RNAfold on commonly used datasets.