Pfold

Pfold predicts RNA secondary structures by integrating evolutionary models and phylogenetic information using stochastic context-free grammars (SCFGs).

Key Features:

• Evolutionary Model Integration: Employs SCFGs to establish prior probability distributions over RNA secondary-structure configurations.
• Phylogenetic Tree Estimation: Estimates phylogenetic trees relating input RNA sequences via maximum likelihood (ML) estimation based on the SCFG model.
• Bayesian Approach: Applies a Bayesian framework using maximum a posteriori (MAP) estimation to combine sequence data with evolutionary priors.
• Performance with Small Sequence Sets: Demonstrates improved prediction accuracy compared to some automated methods, particularly when only a small number of related RNA sequences are available.

Scientific Applications:

• RNA structure-function analysis: Provides secondary-structure predictions to inform studies of RNA functional roles.
• Evolutionary biology and conservation studies: Uses phylogenetic information to detect structurally conserved RNA elements across related sequences.
• Molecular genetics and disease mechanism investigation: Aids exploration of potential impacts of RNA misfolding and structural variation in genetic and disease contexts.

Methodology:

Utilizes SCFGs to model RNA sequences and derive prior probabilities for structures; constructs phylogenetic trees using ML estimation based on the SCFG model; and applies MAP estimation within a Bayesian framework to produce final structure predictions.