Contribute to Rfam
Submit an Alignment
Contribute a Stockholm-format alignment for consideration as a new Rfam family or as an improved seed alignment for an existing family. Submissions are reviewed by our annotators and considered for the next release.
The Rfam database
Rfam is a database of non-coding RNA families. Each family is represented by a hand-curated multiple sequence alignment (the SEED), a consensus secondary structure, and a covariance model built with Infernal. The model is used to search the Rfamseq database for homologues. Rfam covers structural ncRNAs such as riboswitches, ribozymes, miRNAs, snoRNAs, rRNAs, tRNAs, and cis-regulatory elements. It is curated at EMBL-EBI and released under CC0.
Learn more about how Rfam families are built, from seed alignment curation to covariance model construction and iterative searches against Rfamseq, the underlying nucleotide sequence database of complete, non-redundant, representative genomes on which Rfam is based.
SEED sequence alignments
A SEED is a small — from few up to hundreds — hand-curated alignment of representative sequences for a family, annotated with a consensus secondary structure in WUSS notation, and we require Stockholm format. Sequences should be identified as ACCESSION/START-END, where ACCESSION is an ENA, GenBank, or RNAcentral identifier — e.g. AF228364.1/1-30 or URS0000A76344_911092/1-30.
Example:
# STOCKHOLM 1.0 #=GF ID example_RNA AF228364.1/1-30 CGGCAGAUGAUGAU-UUUACUUGGAUUCCC URS0000A76344_911092/1-30 CGGCGGAUGAUGAU-UUUACUUGGGUUCCC #=GC SS_cons <<<<_____________________->>>> //
Before submitting, we recommend running our validator rfam-seed-qc. It performs checks such as sequence-length consistency, duplicates, overlapping or missing coordinates, accession validity against NCBI, SS_cons WUSS syntax, and overlap with existing Rfam families via cmscan.
Submit your alignments
Use the form below to propose a new family or improve an existing one. Provide your details, a supporting reference, the Stockholm file, and any helpful context.