I'm rather new to RNA folding calculations. I want to compute the free
energy of some possible folds of a number of RNA sequences (all of them
rather short, just 22 bases). Unfortunaterly, RNAdraw and the Vienna
routines are giving me different results, and neither of them is the one
that I was expecting. I don't know where the problem lies. My particular
test case is the following sequence:
ugccauucaagaugaaugguac
When I used the Vienna package (linked from my own c++ code) at 37
degrees, it gave me the structure
ugccauuca aga
||||||||
caugguaagu
When I used RNAdraw, at the same temperature, it gave me
ugccauuca aga
||||||||
caugguaagu
So, one program excludes an "a-u" base pair at one end of the stack, and
the other program excludes an "a-u" base pair at the other end. Of
course, I was expecting both base pairs to be included (and hence, a 9
bp stack). When I forced the 9 base pairs on RNAdraw, it gave me a very
high energy (it jumped from approx. -9 Kcal to 800 Kcal).
Since I'm not a biologist (just a mathematician) I don't know if there
is some biological reason for not using the 9 base pairs. But since this
programs are using classical algorithms, which assume the energy of a
base pair to depend only on the immediate neighbours, I don't understand
neither the discrepancy between the two programs nor the discrepancy of
both with my expectations. Any hint? Please!
---
