SPatt

SPatt assesses the statistical significance of DNA patterns within genome sequences by computing P-values under random Markovian background models to identify unusual sequence motifs.

Key Features:

• Pattern Occurrence Counting: The S-SPatt extension counts occurrences of sequence patterns in text files for analysis of large genomic datasets.
• Binomial P-value Approximation: Uses a binomial approximation to compute P-values assuming sequences are generated from a random Markovian source.
• Large Deviation Approximations: Provides large deviation approximations suitable for low-probability patterns for computational efficiency.
• Exact Computations Using Simple Recurrences: Implements exact significance calculations via simple recurrence relations for precise probability estimates.
• Method Selection by Comparison: Compares large deviation approximations and exact recurrence computations to guide selection of the appropriate method.

Scientific Applications:

• Genomic Motif Discovery: Identifying and statistically validating sequence motifs that may regulate gene expression or other genomic functions.
• Comparative Genomics: Detecting significant pattern differences between species or strains to support evolutionary analyses.
• Functional Genomics: Assessing the functional implications of specific DNA sequences by testing motif significance within genomic contexts.

Methodology:

Counts pattern occurrences with S-SPatt; computes P-values via a binomial approximation under a random Markovian source; applies large deviation approximations for low-probability patterns and exact computations using simple recurrence relations, and compares these approaches to select the appropriate method.