Table 1.

Definition of terms used in describing the MEME algorithm

n	number of input sequences
L	length of input sequences
X = {X1, ..., Xn}	the set of n input sequences
w	width of a MEME motif
m = L - w + 1	number of positions for a site
γ	probability of a site in any sequence
θ	PSPM model of motif;
P = {Pi,j}	position-specific prior (PSP)
w0	width for which input PSP is defined
Z = {Zi,j}	missing information variables for i ∈ [1, n],j ∈ [-L, L]
Z(t)	expectation of Z at EM iteration t
= Pr(Zi,j = 1|ϕ(t))	prior probability given PSP & model
ϕ(t)	model parameters at EM iteration t
ϕ = {θ, γ, P}	all sequence model parameters