Of the 16 SARS coronavirus strains submitted to Genbank before June, 2003, 6 had accurate date of host death recorded.
Therefore, this date estimate should be taken as a minimum date of host switch with the actual date possibly being much earlier.
Second, because the date of host death is not good reflection of real time of sequences, the error of time is quite large.
This basically left us with little other choice than to accept the date of host death as the most accurate available estimate for the age of each strain.
Additionally, indirect dating of host molecules is possible in cases such as i) segmental duplications with a differential content of, and ii) haplotypes bearing insertional polymorphism for, 'complete' LTR-elements.
Among the 16 full-length SARS coronavirus genomes, we selected 6 strains for which the accurate date of host death is known, and on which our modelling was based.
As such data were neither included in the submissions to Genbank, nor obtainable by direct contacts to the sequencing labs, we were left to choose between less ideal age estimates for the strains, such as date of host death, sequencing date, or submission date to Genbank.
Based on 6 strains with accurate date of host death, we estimated a time of the last common ancestor, which is coincident with epidemic investigations, and an evolution rate in the same range as that reported for the HIV-1 virus.
For an ancestor sequence S0 of a strain S, we can deduce for the assumptions above that E(D S0, S)) ≈ K(T - T0), where D S0, S) is difference of the two sequences (as depicted by Hamming distance), T0 is the date of the last ancestor, T is the date of host death (as an estimate of sampling date), and K is the evolution rate constant.
The date of the host jump must have occurred between the date of the most common recent ancestor of the Bovidae-infecting clade and the node ancestral to that (i.e, the common ancestor of the Bovidae clade and ST945).
