Infernal

Infernal identifies RNA sequences and secondary structures in DNA sequence databases using covariance models (CMs) to detect RNA homologs that conserve secondary structure over primary sequence.

Key Features:

• Covariance models (CMs): Uses probabilistic profiles (CMs) that capture both sequence consensus and secondary structure consensus of RNA families.
• CM construction: Builds CMs from structurally annotated multiple sequence alignments provided as input.
• Homology search and alignment: Searches large sequence databases for new family members and produces large-scale multiple sequence alignments.
• Filter pipeline (v1.1): Implements a filter pipeline that uses accelerated profile hidden Markov model (HMM) methods and HMM-banded CM alignment techniques.
• Performance improvements: Achieves approximately 100-fold acceleration over the previous version and about 10,000-fold speedup relative to exhaustive non-filtered CM searches.

Scientific Applications:

• RNA homology searches: Detection of homologous RNAs across sequence databases using combined sequence and structure information.
• Discovery of conserved structural motifs: Identification of conserved secondary-structure motifs across diverse RNA families.
• RNA structure–function studies: Facilitates investigations into structure–function relationships of RNAs by finding structurally conserved homologs.
• Large-scale genomics analyses: Enables scalable searches and alignments suitable for large nucleotide datasets in genomics and molecular biology research.

Methodology:

Constructs covariance models from structurally annotated multiple sequence alignments and employs a v1.1 filter pipeline combining accelerated profile HMM methods with HMM-banded CM alignment techniques.