No replies to this topic

[deha]

Hi,
please, does anyone know why I get not calculable "values" for certain pairwise alignments when calculating synonymous and non-synonymous substitutions? (both when using just number of differences and distances, however the proportion of n/c differs between the models used).
I had this problem if just a single codon was aligned, eg. folowing alignment:
1.GAA
2.GAG
3.GTA
4.AAA
5.GTC
6.GAC

If number of differences was calculated using Nei-Gojobori model, then:
[       1     2     3     4     5     6 ]
[1]
[2]  n/c
[3]  n/c   n/c
[4]  n/c   n/c  n/c
[5]  n/c    n/c 1.000 n/c
[6]  n/c    n/c  n/c    n/c  n/c

I could understand that there is a problem for the pairwise align. where there is more than one substitution between the two codons, but the n/c is present even for single-substitution pairs.

Thanks very much if someone can help.
deha