Figure 2

Comprehensive analysis of mutation induced structure disruption in five non-coding RNAs. A.) eSDC (experimental Structure Disruption Coefficient) for 470 single or double mutatants relative to the RNA’s WT sequence. eSDC is computed as one minus the Pearson correlation coefficient of the SHAPE profile (mutant to WT) multiplied by the square root of the length of the RNA. We see that most mutations have small eSDC values indicating that they do not significantly disrupt structure. The five RNAs studied are the human FTL 5’ UTR (FTL), the V. vulnificus Adenine Riboswitch (Adenine RS), the V. cholera Glycine Riboswitch (Glycine RS with and without Glycine (G) bound), the cyclic di-GMP Riboswitch (bis-(3'-5')-cyclic dimeric guanosine monophosphate Riboswitch with and without cyclic-diguanosine-monophosphate (CDM)) and the P4P6 domain of the L-21 Tetrahymena thermopila group I intron [5, 34, 35, 63]. All data were collected under near physiological solution conditions, i.e. 10mM MgCl₂ and 100 mM monovalent. For FTL, hyperferritinemia associated mutations are indicated in magenta. The eSDC values for ± ligand for the three Riboswitches are indicated with a green horizontal line and represent a “biological” threshold above which a structure change is likely to have a functional consequence. This histogram to the right represents a pairwise “within” eSDC calculation for 6-fold repeats of the SHAPE experiments on the FTL UTR RNA to evaluate the reproducibility and significance (p-value) of eSDC values. B.) SHAPE profiles for the WT, U32A and U113A (black, blue, and red respectively) Glycine Riboswitch in the presence of Glycine showing that the U113A mutation very significantly disrupts structure. C.) SHAPE profiles for WT, C128G and C65G (black, blue, and red respectively) P4P6 group I intron transcripts showing that the C65G globally affects the structure of the RNA.