Fig. 5: Two-dimensional mapping of RNA structure in 1G/2G dimer and 3G monomer populations.
From: Short- and long-range interactions in the HIV-1 5′ UTR regulate genome dimerization and packaging
a, Mutations disrupting RNA stems lead to increases in DMS reactivity at positions opposite the mutation. Positions of DMS modification are read out as mutations leading to correlated mutations at pairs of nucleotides involved in RNA structure. RNA secondary structures (blue circles) are identified along the diagonals. Punctate signals (purple circles) can signify noncanonical or tertiary interactions. b,e, The z-score analysis of mutation frequencies from 1G/2G dimer (b) and 3G monomer populations (e). Raw z-scores (lower diagonal) reveal pairs of positions enriched with mutations. Filtered z-scores (upper diagonal) enhance stem signals by applying a convolution filter and signal threshold. Insets are zooms of the filtered z-scores for the polyA–SL1, PBS–SL1 and SL1 stems. c,f, Stem detection in 1G/2G dimer (c) and 3G monomer populations (f). All stems (lower diagonal) reveal all possible stems of minimum length 3 by applying a filter for Watson–Crick and Wobble base pairs to the filtered z-score. Best stem (upper diagonal) selects the best nonconflicting stems by removing conflicting stems based on filtered z-score. Colored boxes represent regions that are highlighted in enhance RNA secondary structure models. d,g, Enhanced RNA secondary structure models of 1G/2G dimer (d) and 3G monomer populations (g). Colored base pairings were detected in multidimensional mapping and used as hard constraints before in RNA secondary structure prediction. Dark blue is TAR. Light blue is polyA. Orange is PBS. Mustard is SL1. Dark green is SL2. Light green is SL3. Red represents the polyA–SL1 interactions, pink shows the new SL1–PBS interaction and purple the TAR–PBS interaction.
