Calculate the Barrier Tree

  $ echo UCCACGGCUGUUAGUGGAUAACGGC | RNAsubopt --noLP -s -e 10 > barseq.sub
  $ barriers -G RNA-noLP --bsize --rates < barseq.sub > barseq.bar

You can restrict the number of local minima using the barriers command-line option -max followed by a number. The option -G RNA-noLP instructs barriers that the input consists of RNA secondary structures without isolated basepairs. -bsize adds size of the gradient basins and -rates tells barriers to compute rates between macro states/basins for use with treekin. Another useful options is -minh to print only minima with a barrier $> dE$. Look at the output file less -S barseq.bar. Use the arrow keys to navigate.

     UCCACGGCUGUUAGUGGAUAACGGC
   1 ......(((((((.....))))))) -7.00  0  9.70   306    0 -7.253634   66 -7.033551
   2 (((((........)))))....... -6.40  1  8.70    49   31 -6.512183   40 -6.512029
   3 ....((..((((....)))).)).. -0.60  2  2.70     6   39 -0.801478   19 -0.824413
   4 (((...(((...))))))....... -0.30  2  0.90     1   10 -0.300000   14 -0.633640
   5 .........................  0.00  2  0.90     3   14 -0.326648  144 -0.634416
   6 ......(((....((.....)))))  0.40  1  0.40     1   21  0.400000    4  0.367865
   7 .((....((......)).....)).  0.50  2  0.50     1   21  0.500000    7  0.300247
   8 ......((((((....)))...)))  0.80  1  1.50     1   82  0.800000    8  0.716052

The first row holds the input sequence, the successive list the local minima ascending in energy. The meaning of the first 5 columns is as follows

1. label (number) of the local minima (1=MFE)
2. structure of the minimum
3. free energy of the minimum
4. label of deeper local minimum the current minimum merges with (note that the MFE has no deeper local minimum to merge with)
5. height of the energy barrier to the local minimum to merge with
6. numbers of structures in the basin we merge with
7. number of basin which we merge to
8. free energy of the basin
9. number of structures in this basin using gradient walk
10. gradient basin (consisting of all structures where gradientwalk ends in the minimum)