RNASUBOPT

NAME

SYNOPSIS

DESCRIPTION

Alternatively, when used with the -p option, RNAsubopt produces Boltzmann weighted samples of secondary structures.

Sequences are read in the usual format, i.e. each sequence occupies a single line, possibly preceded by a fasta-style header line of the form
> name

OPTIONS

-e range

Calculate suboptimal structures within range kcal/mol of the mfe. Default is 1.

-s

Sort the structures by energy. Since the sort in is done in memory, this becomes impractical when the number of structures produced goes into millions. In such cases better pipe the output through `sort +1n'.

-p n

Instead of producing all suboptimals in an energy range, produce a random sample of n suboptimal structures, drawn with probabilities equal to their Boltzmann weights via stochastic backtracking in the partition function. The -e and -p options are mutually exclusive.

-d[0|1|2|3]

Change treatment of dangling ends, as in RNAfold and RNAeval. The default is -d2 (as in partition function folding). If -d1 or -d3 are specified the structures are generated as with -d2 but energies are re-evaluated before printing.

-logML

re-calculate energies of structures using a logarithmic energy function for multi-loops before output. This option does not effect structure generation, only the energies that is printed out. Since logML lowers energies somewhat, some structures may be missing.

-ep prange

Only print structures with energy within prange of the mfe. Useful in conjunction with -logML, -d1 or -d3: while the -e option specifies the range before energies are re-evaluated, -ep specifies the maximum energy after re-evaluation.

-noLP

Only produce structures without lonely pairs (helices of length 1). This reduces the number of structures drastically and should therefore be used for longer sequences and larger energy ranges.

-circ

Assume a circular (instead of linear) RNA molecule.

The -T, -4, -noGU, -noCloseGU, -P, -nsp, options work as in RNAfold.

REFERENCES

VERSION

AUTHORS