PCSCS

PCSCS computes the statistical power of alignment-free k‑tuple sequence comparison statistics D2, D2*, and D2S to evaluate their ability to detect shared motifs and pattern transfers between sequences.

Key Features:

• Alignment-free k‑tuple counting (D2): Counts matching k‑tuples between two sequences without requiring sequence alignment.
• Centralized counts (D2*): Uses centralized counts to modify basic tuple counts for improved signal detection.
• Self-standardization (D2S): Applies a self-standardization approach to tuple counts to increase robustness and interpretability.
• Theoretical power analysis: Assesses theoretical power of D2, D2*, and D2S under different sequence models, including shared‑motif and pattern‑transfer models.
• Numerical simulations and limit distributions: Employs numerical simulations and uses limit distributions of count statistics under null and alternative hypotheses to evaluate performance.
• Simulation benchmarks: Includes simulation studies with sequence lengths up to 140,000 base pairs that demonstrate relative power differences among the statistics, notably D2* often exhibiting superior power.

Scientific Applications:

• Motif detection: Detects shared motifs between sequences, with D2* showing near‑100% power when a sufficient number of motifs are shared.
• Pattern transfer analysis: Evaluates scenarios of pattern transfer where increasing sequence length does not increase statistical power for D2, D2*, or D2S.
• Conserved element and evolutionary analysis: Assesses detection of conserved elements such as transcription factor binding motifs and supports studies of genomic conservation and evolutionary relationships.

Methodology:

Performs theoretical examination using limit distributions of count statistics under null and alternative hypotheses and conducts numerical simulations (including sequences up to 140,000 base pairs) to determine the power of D2, D2*, and D2S under specified models.