sarks

sarks identifies sequence motifs whose presence correlates with numeric scores assigned to sequences such as gene promoters.

Key Features:

• Suffix array kernel smoothing: Smooths sequence scores over sequence similarity quantified by location within a suffix array based on the full set of input sequences.
• Multi-motif domain detection: A second round of smoothing over spatial proximity within sequences reveals multi-motif domains (MMDs).
• Motif merging and extension: Discovered motif sites can be merged or extended based on adjacency within MMDs.
• False positive control: False positive rates are estimated and controlled by permutation testing.
• Non-parametric approach: Applies non-parametric kernel smoothing to uncover promoter motif sites that correlate with elevated differential expression scores.

Scientific Applications:

• Differential gene expression analysis: Mines differential gene expression data to identify candidate cell type-specific DNA regulatory sequences using continuous scores such as fold-change, test-statistic, or P-value.
• Neocortical neuron class characterization: Applied to published gene expression data representing distinct neocortical neuron classes in Mus musculus.
• Interneuron developmental state analysis: Applied to interneuron developmental states in Homo sapiens.

Methodology:

sarks detects motif k-mers by smoothing sequence scores over sequence similarity via suffix array kernel smoothing, followed by a second round of smoothing over spatial proximity to reveal multi-motif domains. Discovered motif sites are merged or extended based on adjacency within MMDs, with false positive rates controlled by permutation testing.