Fig. 1

a Minimum free energy (MFE) structure of the initial 52-nt Gag-Pol overlapping reading frame in positions 1631-1682 of the HIV-1 complete genome (GenBank AF033819.3). This frameshift stimulating signal (FSS) contains the initial slippery sequence heptamer, given by U UUU UUA in the Gag reading frame, as well as the displayed stem-loop secondary structure, which together promote a programmed -1 frameshift UUU UUU A in the Pol reading frame. b Depiction of all 6 possible reading frames – RNAsampleCDS samples RNA sequences that code in all possible reading frames, allowing IUPAC sequence constraints c Sequence logo for 145 RNA HIV-1 frameshift signal sequences from the RF00480 seed alignment from Rfam 12.0 [4]. d Sequence logo for the Pol peptide coded by 138 RNA HIV-1 frameshift signal sequences from the RF00480 seed alignment from Rfam 12.0; Pol peptide translated from nucleotide positions 1-51. e Sequence logo for the Gag peptide coded by 138 RNA HIV-1 frameshift signal sequences from the RF00480 seed alignment from Rfam 12.0; Gag peptide translated from nucleotide positions 2-52. Since some sequences from RF00480 contained IUPAC codes for uncertain data, the data were disambiguated–for instance, the code B (not A) was disambiguated by randomly assigning either C,G or U with probability 1/3. Seven sequences were removed from the seed alignment of 145 RNAs due to gaps in the alignment, and another five sequences were removed since either the Pol or Gag peptide contained a stop codon–resulting in 133 sequences for nucleotide analysis. Peptide sequence logos for the 138 Pol and Gag peptides were created using WebLogo [26]