For determining the motifs in a DNA sequence, methods such as Position Weight Matrix are being used. Recently first order Markov models are also being used for such purposes. Usefulness of these models for training data has been well established. Analysis of motif search has been restricted by the fact as to how the variants of the site would affect the site.
Recently Korhonen J.H. (2017) have developed an algorithm for searching motifs in the sequences that majorly overcomes the drawbacks of existing motif search algorithm. Their method is based on two important aspects (1) How to use PWM matching algorithms for high order of Position Weight Matrix framework. (2)How the indels, SNPs and other sequence analysis variants can be used for developing motif search algorithm.
Implementation of the algorithm are available at the following website : http://www.cs.helsinki.fi/group/pssmfind
Reference:
Korhonen J.H. et al. (2017)Fast motif matching revisited: high order PWMs, SNPs and indels. Bioinformatics 33(4): 514-521.