algorithm algorithm Cosmo cosmo allows the user to target the motif search by specifying a set of constraints that the unknown position weight matrix must satisfy. The algorithm is based on a probabilistic model that describes the DNA sequences of interest through a two- component multinomial mixture model with estimates of the position weight matrix entries obtained by maximizing the observed data likelihood over the smaller parameter space corresponding to the imposed constraints. It includes methods such as Probabilistic models and one-occurrence-per-sequence and zero-or-one-occurrence-per-sequence and two-component mixture