PyScale — Nucleotide Propensity-Scale Toolkit
- bioinformatics
- Python
- dynamical systems
- genomics
What it is
A bioinformatics toolkit that turns nucleotide sequences into numerical propensity profiles for alignment-free analysis.
It ships 267 curated propensity scales (mono-, di-, and trinucleotide), four profiling paradigms (window averaging, periodicity, flip-reading, complementary strand), and a robust largest-Lyapunov-exponent estimator (enhanced Rosenstein algorithm with k-NN averaging and automatic fit-region selection) for sequence-complexity estimation — validated against the chaotic logistic map. A dedicated promoter analyzer detects TATA motifs and computes GC content, demonstrated on human TATA-box vs. CpG-island promoters. Ships with CLI and GUI tools. (Co-authored.)
Stack
Python · NumPy · matplotlib · Tkinter (GUI) · Rosenstein Lyapunov estimation