Having introduced various methodologies for modeling sequences, the focus now shifts to describing statistical modeling procedures for DNA patterns. There is a growing recognition of numerous well-established patterns that are essential for understanding and analyzing DNA sequences. These patterns, often of functional significance, are typically discovered through multiple-sequence alignments of related sequences within a specific family. During the alignment process, sequences may exhibit gaps of varying sizes, but certain regions remain consistently aligned without gaps across all sequences. These fixed-size, ungapped, aligned regions represent the patterns that require precise modeling to facilitate the identification of similar patterns within anonymous DNA segments. One effective statistical technique for this purpose is based on Hidden Markov Models (HMMs), which can be employed to develop a closed-form representation of these patterns.