Dependency of codon usage on protein sequence patterns: a statistical study

Table 4 Mutual information between “pattern gene codon usage (RCU)” and “pattern region codon usage (GCU)”

Amino acid	Number of codons of amino acid	Number of valid patterns	Number of patterns with mutual information less than 0.0001 bits	Percentages of patterns with mutual information less than 0.0001 bits	Average of mutual information of the patterns
C	2	561	491	%87.5	0.00009
D	2	809	735	%90.9	0.00007
E	2	772	693	%89.8	0.00005
F	2	761	686	%90.1	0.00007
H	2	609	525	%86.2	0.00007
K	2	764	689	%90.2	0.00006
N	2	724	660	%91.2	0.00005
Q	2	639	533	%83.4	0.00009
Y	2	668	600	%89.8	0.00007
I	3	860	715	%83.1	0.00008
A	4	871	646	%74.2	0.00015
G	4	937	659	%70.3	0.00018
P	4	688	454	%66.0	0.00023
T	4	824	601	%72.9	0.00016
V	4	894	683	%76.4	0.00012
L	6	831	499	%60.0	0.00026
R	6	650	355	%54.6	0.00035
S	6	724	491	%67.8	0.00017

The mutual information is computed between two random variables, namely, “pattern gene codon usage (GCU)”, and “pattern region codon usage (RCU)”. We excluded amino acids with exactly one codon. If an amino acid appears 30 times in a pattern region, and each of its codons appears at least once in this region, we consider the pattern as a valid pattern for the amino acid.

ISSN: 1742-4682