Comparison of normalization methods for Hi-C data.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• High-throughput chromosome conformation capture (Hi-C) is crucial for studying genome-wide interactions.
• Systematic biases in raw Hi-C data introduce variability, impacting downstream analysis reliability.
• Normalization is essential for mitigating these biases in Hi-C data processing.

Purpose of the Study:

• To comprehensively compare six different Hi-C normalization methods.
• To evaluate method performance across multiple considerations for bias reduction.
• To provide practical recommendations for selecting appropriate Hi-C normalization strategies.

Main Methods:

• Investigated six distinct Hi-C normalization algorithms.
• Performed a comparative analysis based on predefined criteria.
• Evaluated both individual sample and cross-sample normalization approaches.

Main Results:

• Cross-sample normalization methods demonstrated superior performance compared to individual sample methods in most evaluated aspects.
• Identified key differences in the effectiveness and underlying mechanisms of the compared normalization techniques.
• Results are summarized in a comparative table to guide method selection.

Conclusions:

• Cross-sample normalization is recommended for enhancing Hi-C data quality and downstream analysis accuracy.
• The study provides valuable insights for researchers to choose optimal Hi-C normalization pipelines.
• Source code for implemented methods is publicly available for reproducibility and further research.