mfold

Mfold is a computational tool for predicting the secondary structure of single-stranded nucleic acids. It was developed to estimate how an RNA or DNA sequence folds by minimizing free energy under defined conditions.

The program applies dynamic programming to enumerate possible base-pairings and finds the conformation with the lowest free energy using thermodynamic parameters derived from nearest-neighbor models (Turner parameters). It can also report several suboptimal structures within a specified energy range. The input is a nucleotide sequence; the user may specify temperature and ionic conditions.

Output includes the predicted secondary structure in dot-bracket notation, a corresponding base-pairing diagram or energy graph, and the free-energy value of the predicted structure.

Historically, Mfold has been widely used in RNA biology and bioinformatics; it exists as a standalone tool and as a web server in various versions. While it approximates nucleic acid folding and cannot model pseudoknots in its standard form, and its energy models have been refined since, it remains a foundational reference in secondary structure prediction.