Table 1.

Sets of residues unrelated to the crosstalk between phosphorylation and sulfoxidation do not show coevolutionary signal

Group	Set at 218	Set at 222	LRT	p-value
Positive	{S,T}	{M}	15.9	0.003
Control-1	{S,T}	{V}	2.4	0.668
Control-1	{S,T}	{C}	3.1	0.542
Control-1	{S,T}	{N}	4.6	0.336
Control-1	{S,T}	{L}	0.0	1.000
Control-2	{N,Q}	{M}	3.7	0.449
Control-2	{G,A}	{M}	0.0	1.000
Control-2	{D,E}	{M}	0.0	1.000
Control-2	{H,K}	{M}	0.0	1.000
Control-3	{Q,N}	{V}	0.2	0.993
Control-3	{G,A}	{V}	6.2	0.181
Control-3	{D,E}	{V}	0.0	1.000
Control-3	{H,K}	{V}	3.5	0.475
Control-4	{Q,P}	{G}	0.0	1.000
Control-4	{W,V}	{A}	0.0	1.000
Control-4	{L,T}	{I}	0.0	1.000
Control-4	{I,Y}	{H}	0.0	1.000

The existence of correlated evolution between the sites 218 and 222 was examined using different sets of residues to define the model states. Thus, beside the sets formed by phosphorylatable residues at 218 and sulfoxidable methionine at 222, used as positive reference, four other types of set combinations were analyzed. In the first type (Control-1, rows 2–5), the residues providing X = 1 are still serine and threonine, but the residue making Y = 1 is now different to methionine (as indicated in the table). Val, Cys, Asn and Leu have been selected because of their frequencies (from higher to lower) at position 222 in the eukaryotic species examined. The second group of control analyses (Control-2, rows 6–9) always included methionine at position 222 as the residue providing Y = 1, while the set leading to X = 1 was now formed by a pair of non-phosphoaceptor amino acids of similar physicochemical properties. The third group (Control-3, rows 10–13) was formed by sets of residues that were neither phosphorylatable nor sulfoxidable. Finally, the fourth group (Control-4, last four rows), was established by randomly taking the sets, among the twenty proteinogenic amino acids. The LRTs obtained, and their respective p-values, are shown