Chronolab: an interactive software package for chronobiologic time series analysis written for the Macintosh...

Area of Science:

• Chronobiology
• Biostatistics
• Computational Biology

Background:

• Short, noisy, and nonequidistant time series are common in medicine and biology.
• Detecting periodic components in such data requires specialized methods.
• Periodic regression offers a framework for analyzing time series with inherent variability.

Purpose of the Study:

• To describe an interactive least-squares rhythmometry program for Macintosh computers.
• To provide tools for analyzing periodic components in biological and medical time series.
• To facilitate the detection and characterization of biological rhythms.

Main Methods:

• Development of an interactive C-language program for least-squares rhythmometry.
• Implementation of sequential fitting of trial periods (linear in time) and harmonic components (linear in frequency).
• Inclusion of data transformations and statistical tests (e.g., p-value for zero amplitude, sinusoidality, normality, homogeneity of variance).

Main Results:

• The program offers detailed output including fitted period, percent rhythm, p-values, mesor, amplitude, acrophase, standard errors, and confidence intervals.
• It provides summary statistics and allows for multiple-component and comparative analyses across individuals and variables.
• User-friendly menus and self-explanatory commands facilitate data analysis.

Conclusions:

• The developed program is a powerful tool for rhythmometric analysis of complex biological time series.
• It enhances the ability to detect and quantify periodic phenomena in medical and biological research.
• The software supports simultaneous analysis of multiple series and individuals, aiding in comparative chronobiology studies.