DiscoRhythm

DiscoRhythm detects and characterizes rhythmic signals in high-throughput time-series -omics datasets by estimating rhythmic parameters such as phase, amplitude, and statistical significance.

Key Features:

• Rhythm detection algorithms: Implements Cosinor, JTK Cycle, ARSER, and Lomb-Scargle algorithms for detecting oscillatory signals.
• Rhythmic parameter estimation: Estimates phase, amplitude, and statistical significance for detected rhythms.
• Computational optimization: Optimized implementations yield execution-time improvements reported up to 30-fold for large-scale datasets.
• Dimensionality reduction: Provides methods for dimensionality reduction tailored to rhythmic data analysis.
• Periodicity profiling: Supports periodicity profiling across features to characterize rhythmic behavior.
• Experimental replicates: Incorporates experimental replicates into rhythmic analyses.
• R/Bioconductor implementation: Implemented as an R/Bioconductor package for use within R-based analysis workflows.

Scientific Applications:

• Chronobiology: Identification and characterization of oscillating genes, proteins, or metabolites across circadian and other biological rhythms.
• Systems biology: Profiling rhythmic behavior at the network or pathway level to study temporal coordination of biological systems.
• Medical research: Detection of rhythmic biomarkers and investigation of temporal alterations relevant to disease mechanisms and therapeutic targeting.
• High-throughput -omics studies: Scalable analysis of -omic-scale time-series datasets to enable rapid screening for oscillatory signals.

Methodology:

Applies Cosinor, JTK Cycle, ARSER, and Lomb-Scargle algorithms to estimate phase, amplitude, and statistical significance; includes dimensionality reduction, periodicity profiling, and handling of experimental replicates, with optimized implementations reported to improve execution time up to 30-fold.