A Bootstrap Based Measure Robust to the Choice of Normalization Methods for Detecting Rhythmic Features in High...

Area of Science:

• Genomics
• Systems Biology
• Bioinformatics

Background:

• High-throughput gene-expression data require pre-processing, including normalization, to remove noise and systematic variations.
• The selection of normalization methods can significantly influence the identification of rhythmic genes, particularly in oscillatory systems like the circadian clock.
• Accurate identification of rhythmic genes is crucial for toxicological and pharmacological studies.

Purpose of the Study:

• To develop a rhythmicity measure and bootstrap methodology for detecting rhythmic genes robustly.
• To assess the impact of different normalization methods on gene rhythmicity detection.
• To ensure that identified rhythmic genes are not artifacts of the chosen normalization technique.

Main Methods:

• Introduction of a novel rhythmicity measure.
• Application of a bootstrap methodology for robust gene detection.
• Validation using publicly available circadian clock microarray gene-expression datasets.

Main Results:

• The proposed methodology demonstrates high correlation in rhythmicity measure values across different normalization methods.
• Gene rhythmicity detection is shown to be robust and minimally affected by the choice of normalization technique.
• The bootstrap methodology can also be utilized for simulating gene expression data in oscillatory systems.

Conclusions:

• The developed rhythmicity measure and bootstrap methodology provide a robust approach for identifying truly rhythmic genes.
• The findings suggest that gene rhythmicity is less likely to be an artifact of normalization when using this method.
• The methodology offers a reliable tool for analyzing gene expression data in oscillatory biological systems.