Transition/Transversion Stepmatrices

Question

[Yongjie Zhang]

Hi all,

I want to set transition/transversion stepmatrices during doing phylogenetic analysis. Can anyone tell me how to make it? Is there an easier way than manual counting to count the number of transition sites and the number of transversion sites in a locus alignment? Then how to set up the stepmatrices based on those two numbers if their ratio is not integer?

Thanks.

Yongjie

[Biomonika (Noolean)]

Could you please list example of what you are trying to achieve? What is the problem? How do the sequences look like?

[Yongjie Zhang]

Hi Noolean,

Let me take my ITS alignment as example. Among the total 543 sites in the alignment, there are 36 parsimony informative sites, 5 parsimony uninformative sites and 6 gap sites. Of the 36 informative sites are 16 T/C & 9 A/G transitions, 3 A/C, 2 T/G, 3 C/G & 1 A/T transiversions, and 2 sites each with 3 variants (e.g, A/T/C at one site), My questions actually includes:

1) If there is a program or script that can automatically count the numbers of transition and transversion sites existed in a alignment? It found it a hard job if I have many variable sites to count by eye and it is prone to error.

2) When I establish the stepmatrix, should I consider both informative sites and uninformative sites? How about the sites with 3 variants?

3) In this example, the ratio of transition to transversion is 2.56 when only informative sites are considered. What value should I use to set up the stepmatrix. What I have read in literature are always integers, like

usertype 2_1 stepmatrix = 4 acgt

'- 2 1 2 2( ' is added in the post to correctly display "-")

2 - 2 1

1 2 - 2

2 1 2 -

4) If I don't want to miss the information of gaps, how to set up the stepmatrix? What value can be given to gap in the stepmatrix? How about the following stepmatrix although X represent some value given to gap?

usertype 2_1 stepmatrix = 4 acgt-

0 2 1 2 X

2 0 2 1 X

1 2 0 2 X

2 1 2 0 X

X X X X 0

I found when I define stepmatrix for variable sites including not gaps and at the same time set gapmode to fifth base, there are always conflict and my PAUP program stops as I posted at How to define gap cost in a stepmatrix?.

Hope to have your instruction.

Thanks.

Yongjie