PMC

Topology of the predicted consensus secondary structure of the HIV-1 genome, predicted with PPfold 3.1. Base pairing is indicated by blue lines. The structure was drawn in jViz.Rna () and annotated manually after visual inspection.

Putative long-distance interactions in the HIV-1 genome. (A) Core organizing structure (COS) in the central domain of the HIV-1 genome, coordinating the formation of three arms. (B) Putative long-distance interactions in the ‘RRE domain’ of the HIV-1 genome. The letters correspond to those in Figure . Base pairs with significant base pair covariations are indicated, in the 5′-3′ direction.

Phylogenetic analysis of the proposed structural model. The proportion of base pair covariations (two-sided covariation) is plotted. This is the proportion of base pairs consistent with the structural prediction (A-U, G-C, G-U), when removing gaps and the most frequent canonical base pair, as described in the Materials and Methods section. A higher resolution version of the figure with nucleotide identities is available in Supplementary Figure S4.

The PPfold 3.1 structure coloured by comparison with the Watts09 structure. The predictions were made with identical SHAPE data. Common structural elements are indicated in red (base pairs) and blue (unpaired nucleotides). Elements paired in both structures (but with different partners) are indicated by yellow. Nucleotides paired in one but unpaired in the other are indicated in brown. 72% of all unpaired nucleotides and 31% of base paired nucleotides were found in both predictions. Over half of the genome is unpaired in both predictions (blue). A higher resolution version of the figure with nucleotide identities is available in Supplementary Figure S3.

A comparative model for the secondary structure of the HIV-1 genome, predicted with PPfold 3.1, integrating covariation information from a manually curated alignment of 38 HIV-1 genome sequences and structure probing data from the SHAPE method. Each nucleotide is coloured according to the reliability score (described in Materials and Methods). Long-distance interactions (further than 600 nucleotides apart) are indicated with letter codes. Corresponding capital and non-capital letters interact. A higher resolution version of the figure with nucleotide identities is available in Supplementary Figure S2.

Short-distance interactions supported both by consistency of prediction and covariation. Base pair covariations are indicated as boxes for qualifying base pairs. SHAPE values are indicated on a scale from black (low degree of flexibility) to red (high degree of flexibility). Missing SHAPE values at the start of the 5′ TAR hairpin are indicated in grey. The part of the RRE structure that did not pass the stringent filtering criteria is indicated in light blue. The structures that were not present in the most recent update of the Watts09 structure () are indicated with asterisk (*). Structure 9 is partially present in ().