4
Files included in the folder:
6
mtCDNApri.nuc: codon sequences from 7 apes (small data set in Yang et al. 1998)
7
mtCDNApri.trees: tree file for the data
8
OmegaAA.dat: file specifying different types of amino acid substitutions
9
codeml.ctl: control file
11
Also the following two files are incldued (large data set in Yang et al. 1998)
15
This directory contains example files for estimating
16
nonsynonymous/synonymous substitution rate ratios for different types
17
of amino acid changes. The data were used in Yang et al. (1998), so
18
you can use the data to duplicate results in that paper.
20
(1) Table 5 and the section titled "Different types of amino acid
21
substitutions". This model assumes different dn/ds (w) ratios for
22
different types of amino acid substitutions. You specify how many
23
amino acid substitution types you would like to have and which amino
24
acid pairs (changes) should be in each type. The data should be codon
25
sequences (seqtype = 1), and the model is specified by aaDist = 7
26
(AAClasses). The details of amino acid substitution types are
27
specified in a file called OmegaAA.dat. See that file for details.
28
The model can be used to fit different rates for "radical" and
29
"conserved" amino acid substitutions.
31
Yang et al. (1998) implemented the model for codon sequences only
32
(seqtype = 1). In theory the model should be applicable to amino acid
33
sequences as well (seqtype = 2), with one fewer parameter required.
34
However, such models (for amino acid sequences) are either not tested
35
properly or never made to work. If you need to apply such models to
36
amino acid sequences, you can let me know and I'll try to savalge the model.
39
(2) Table 4. Mechanistic models of codon substitution. You should have
40
seqtype = 1, model = 0, NSsites = 0. Then the models are specified
41
using aaDist as follows. You need to copy files with names like
42
g1974c.dat, g1974p.dat, g1974v.dat, g1974a.dat, grantham.dat, or
43
miyata.dat into the current folder to run those models.
45
aaDist = 1 * geometric relationship using Grantham
46
aaDist = 2 * geometric relationship using Miyata & Yasunaga
47
aaDist = 3 * geometric relationship using c (composition)
48
aaDist = 4 * geometric relationship using p (polarity)
49
aaDist = 5 * geometric relationship using v (volume)
50
aaDist = 6 * geometric relationship using a (aromaticity)
52
aaDist = -1 * linear relationship using Grantham
53
aaDist = -2 * linear relationship using Miyata & Yasunaga
54
aaDist = -3 * linear relationship using c (composition)
55
aaDist = -4 * linear relationship using p (polarity)
56
aaDist = -5 * linear relationship using v (volume)
57
aaDist = -6 * linear relationship using a (aromaticity)
60
(3) Table 3. Mechanistic models of amino acid substitution. As above
61
for table 4, but you should have seqtype = 2, model = 6 (FromCodon),
62
NSsites = 0. Note that with amino acid sequences only, we cannot
63
estimate synonymous rate, and the models have one fewer parameter than
64
when codon sequences are used.
66
The models were implemented several years ago and not carefully
67
maintained since then. So let me know you notice anything strange.
68
You should always run the example data set to duplicate the results in
73
Yang, Z., R. Nielsen, and M. Hasegawa, 1998. Models of amino acid
74
substitution and applications to mitochondrial protein
75
evolution. Mol. Biol. Evol. 15: 1600-1611.