Infernal cmscan (EBI)

Infernal cmscan (EBI) identifies RNA sequences and secondary-structure homologs by searching DNA sequence databases, such as Rfam, using covariance models that capture sequence and secondary-structure conservation.

Key Features:

• Covariance models (CMs): Probabilistic profiles that encode both sequence and secondary-structure characteristics of RNA families derived from structurally annotated multiple sequence alignments.
• HMM-accelerated filter pipeline: Version 1.1 employs accelerated profile hidden Markov model methods to prefilter candidate sequences before CM alignment.
• HMM-banded CM alignment: Uses HMM-banded constraints to restrict the CM dynamic programming search space during alignment.
• Speed improvements: The filter pipeline provides approximately 100-fold acceleration over previous versions and about 10,000-fold improvement compared with exhaustive non-filtered CM searches.
• Search targets: Detects RNA structures and sequence similarities against databases such as Rfam.

Scientific Applications:

• RNA homology search: Detection of homologous RNAs in nucleotide sequence databases using combined sequence and secondary-structure models.
• RNA family member discovery: Identification of new members of known RNA families by matching sequences to covariance models derived from annotated alignments.
• Structural RNA annotation: Annotation of RNA secondary structures and family assignments in genomic and metagenomic sequence data.

Methodology:

Infernal uses covariance models derived from structurally annotated multiple sequence alignments, applies an HMM-accelerated filter pipeline and HMM-banded CM alignment, and searches sequence databases such as Rfam for sequence and structure similarities.