A codon-based model designed to describe lentiviral evolution

Anne Mette K. Pedersen*, Carsten Wiuf, Freddy B. Christiansen

*Corresponding author for this work
32 Citations (Scopus)

Abstract

A codon-based model designed to describe lentiviral evolution is developed. The model incorporates unequal base compositions in the three codon positions and selection against the CpG dinucleotide within codons to account for a deficit of this dinucleotide exhibited by lentiviral genes. The model is, to a large extent, able to account for the pattern of codon usage exhibited by the HIV1 genes gag, pol, and env, in spite of its parameter paucity. The model is extended to a similar model which operates on pentets (codons and their neighboring bases). The results obtained by the pentet model establish the importance of depression of CpGs across codon boundaries as well as within codons. The goodness of fit of the CpG depression model to the observed evolution in pairwise alignments of HIV1 sequences is assessed. The model provides a significantly better description of the observed evolution than the simpler models examined. The parameter estimates indicate that part of the unusually large biases in nucleotide frequencies observed in HIV1 genes is caused by selection against CpGs. We find that the estimates of expected numbers of substitutions, of transitions to transversions, and of synonymous to nonsynonymous substitution rates are robust to CpG depression, whereas the ratio of CpG-generating substitutions to other substitutions is strongly influenced by the choice of model.

Original languageEnglish
JournalMOLECULAR BIOLOGY AND EVOLUTION
Volume15
Issue number8
Pages (from-to)1069-1081
Number of pages13
ISSN0737-4038
DOIs
Publication statusPublished - 1 Jan 1998
Externally publishedYes

Keywords

  • Codon usage
  • Codon-based model
  • CpG depression
  • Goodness of fit
  • HIV1
  • Lentivirus

Fingerprint

Dive into the research topics of 'A codon-based model designed to describe lentiviral evolution'. Together they form a unique fingerprint.

Cite this