cofold

Cofold, short for co-folding, refers to the computational practice of predicting the joint secondary structure of two RNA molecules by considering both their intramolecular base pairing and potential intermolecular base pairing between the two sequences. The approach extends single-sequence RNA folding, which seeks the minimum free energy structure for one molecule, to two sequences in order to model how they may interact and form a complex.

In a typical cofolding analysis, the objective is to determine whether the two RNAs will bind and, if so, which regions are involved in the interaction. The output usually includes a minimum free energy (MFE) structure for the combined system and information about base-pairing probabilities, indicating which nucleotides are likely to form inter- versus intramolecular bonds. The results help identify potential binding sites, docked motifs, or regulatory interactions between the RNAs.

Cofolding is implemented in several software tools, with RNAcofold being a well-known example in the ViennaRNA package. Such tools accept two RNA sequences as input and return the joint MFE structure, along with a dot-bracket encoding that represents both intra- and intermolecular pairings. Other platforms supporting multi-strand or co-fold analyses include various nucleic acid folding suites and web servers.

Applications of cofolding include studying RNA–RNA interactions such as antisense regulation, microRNA target engagement, ribozyme–substrate interactions, and viral or bacterial RNA interactions. Limitations include the common exclusion of pseudoknots in standard cofolding models and reliance on thermodynamic parameterizations, which may not capture all real-world dynamics; experimental validation is often required.