MARS (RNAcmap3)

RNAcmap3 facilitates improved homology search and multiple sequence alignment for non-coding RNAs through integration with the MARS database and a split-search strategy.

Key Features:

• Expanded sequence database: MARS integrates ncRNA sequences from RNAcentral, transcriptome and metagenome assemblies from MG-RAST, genomic sequences from Genome Warehouse (GWH), genomic sequences from MGnify, and NCBI's nt database and its subsets, resulting in a database 20-fold larger than NCBI's nt database or 60-fold larger than RNAcentral.
• Split-search strategy: A new split-search strategy enables substantial improvement in homology search over existing state-of-the-art techniques.
• Multiple sequence alignment generation: RNAcmap3 produces more accurate and more sensitive multiple sequence alignments (MSAs) than manually curated MSAs from Rfam for the majority of structured RNAs mapped to Rfam.
• Fully automatic operation: The homology search process operates without manual intervention.

Scientific Applications:

• Structural inference of ncRNAs: MARS coupled with RNAcmap3 enables improved structural prediction of non-coding RNAs through enhanced homology detection.
• Functional inference of ncRNAs: The integrated system supports improved functional annotation of non-coding RNAs based on detected homologous relationships.
• RNA language model training: The generated MSAs provide training data for RNA language models.

Methodology:

RNAcmap3 employs a split-search strategy against the MARS database, which incorporates sequences from RNAcentral, MG-RAST, Genome Warehouse, MGnify, and NCBI nt.