Table 1.

Ancestral repeats - First-order root sequence distribution

	Root
X	πX|A	πX|C	πX|G	πX|T
A	0.3114 [0.3063; 0.3162]	0.1682 [0.1641; 0.1723]	0.2355 [0.2307; 0.2400]	0.2849 [0.2799; 0.2899]
C	0.3760 [0.3697; 0.3826]	0.2439 [0.2380; 0.2496]	0.0069 [0.0057; 0.0083]	0.3733 [0.3668; 0.3799]
G	0.3014 [0.2954; 0.3077]	0.1926 [0.1873; 0.1981]	0.2448 [0.2389; 0.2504]	0.2612 [0.2554; 0.2672]
T	0.2287 [0.2241; 0.2331]	0.1966 [0.1924; 0.2012]	0.2451 [0.2405; 0.2497]	0.3296 [0.3247; 0.3347]
	Rest of the tree
	πA	πC	πG	πT
	0.3114 [0.3015; 0.3217]	0.1676 [0.1610; 0.1739]	0.1730 [0.1665; 0.1798]	0.3480 [0.3374; 0.3582]

Estimates (mean and accompanying 95% credibility interval) for the four sets of base frequencies at the ancestral root sequence and the set of base frequencies used in the remainder of the tree, for the context-dependent model with independent model frequencies and a first-order Markov chain at the ancestral root. The extremely low estimate for the probability of observing a G when the preceding base is a C, is a direct consequence of the CpG-methylation-deamination process in mammalian sequences. Note that the base frequencies used throughout the remainder of the tree differ from those when assuming a zero order Markov chain at the ancestral root sequence.