Energy minimization methods applied to riboswitches: a perspective and challenges.

Energy minimization methods for RNA secondary structure prediction have been used extensively for studying a variety of biological systems. Here, we demonstrate their applicability in riboswitch studies, exemplified in both the expression platform and aptamer domains. In the expression platform domain, energy minimization methods can be used to predict in silico a unique point mutation positioned in the non-conserved region of the TPP riboswitch that will transform it from a termination to an anti-termination state, thus backing the prediction experimentally. Furthermore, a successive prediction can be made for a compensatory mutation that is positioned over half the sequence length of the riboswitch from the original mutation and that completely overturns the anti-termination effect of the original mutation. This approach can be used to computationally predict rational modifications in riboswitches for both research and practical applications. In the aptamer domain, energy minimization methods can be used when attempting to detect a novel purine riboswitch in eukaryotes based on the consensus sequence and structure of the bacterial guanine binding aptamer. In the process, some interesting candidates are identified, and although they are attractive enough to be tested experimentally, they are not detectable by sequence based methods alone. These brief examples represent the important lessons to be learned as to the strengths and limitations of energy minimization methods. In light of our growing knowledge in the energy minimization field, future challenges can be advanced for the rational design of known riboswitches and the detection of novel riboswitches. Unlike analyses of specific cases, it is stressed that all the results described here are predictive in scope with direct applicability and an attempt to validate the predictions experimentally.