Secondary Structure Ensemble Analysis via Community Detection

We explored the extent to which graph algorithms for community detection can improve the mining of structural information from the predicted Boltzmann/Gibbs ensemble for the biological objects known as RNA secondary structures. As described, a new computational pipeline was developed, implemented, and tested against the prior method RNAStructProfiling. Since the new approach was judged to provide more structural information in 75% of the test cases, this proof-of-principle analysis supports efforts to improve the data mining of RNA secondary structure ensembles.