POXO

POXO discovers regulatory cis-elements by identifying and statistically validating over-represented nucleotide motifs in upstream promoter sequences of co-expressed gene sets to link transcription factor binding sites with gene co-regulation.

Key Features:

• Functional evaluation and grouping: Evaluates biological functions of co-expressed gene sets and groups genes into functionally coherent subsets when multiple processes are present.
• Sequence retrieval and pattern discovery: Retrieves upstream promoter sequences and performs explicit enumeration of nucleotide patterns on the alphabet A, C, G, T, N to identify over-represented motifs.
• Pattern verification: Clusters similar candidate motifs and verifies their significance using bootstrapping to estimate mean occurrences and standard deviations, followed by statistical testing including ANOVA F-statistics, Tukey's honestly significant difference test, and P-values.
• Tailored analytical pipelines: Supports construction of customizable analytical workflows for motif discovery, clustering, and statistical validation.

Scientific Applications:

• High-throughput genomic analyses: Analyzes large gene lists to detect functional heterogeneity and shared cis-regulatory elements among co-expressed genes.
• Functional clustering and interpretation: Applies Non-negative Matrix Factorization (NMF) to cluster genes into homogeneous functional groups to aid interpretation of complex expression data.
• Plant defense and transcription factor discovery: Applied to pathogen-treated Arabidopsis thaliana expression data to identify activated gene sets and suggest candidate transcription factors responsible for regulation.

Methodology:

Retrieves upstream promoter sequences; performs explicit enumeration of nucleotide patterns on the alphabet A, C, G, T, N; uses bootstrapping to estimate mean occurrences and standard deviations; clusters similar motifs; applies statistical validation including ANOVA F-statistics, Tukey's HSD and P-values; applies Non-negative Matrix Factorization (NMF) for functional clustering.