Systematic Evaluation of Statistical Methods for Identifying Looping Interactions in 5C Data.

Area of Science:

• Genomics and Molecular Biology
• Computational Biology and Bioinformatics

Background:

• Chromosome-Conformation-Capture-Carbon-Copy (5C) is a proximity ligation technique for high-resolution analysis of genome-wide looping interactions.
• Existing 5C data analysis pipelines involve complex, interdependent normalization and statistical steps that significantly impact biological interpretations.
• A systematic comparison of method performance across the entire 5C analysis workflow has been lacking.

Purpose of the Study:

• To provide a comprehensive comparative assessment of different computational methods used at each stage of the 5C data analysis pipeline.
• To identify and discuss the trade-offs associated with various analytical approaches.
• To introduce a versatile algorithmic suite, lib5C, enabling researchers to evaluate diverse analytical strategies on their own 5C datasets.

Main Methods:

• Comparative evaluation of algorithms for sequencing depth and library complexity correction.
• Assessment of bias mitigation and spatial noise reduction techniques.
• Analysis of methods for estimating distance-dependent expected values and variances, statistical modeling, and loop detection.

Main Results:

• Detailed comparison of method performance across all key steps of 5C data processing.
• Identification of advantages and disadvantages for each analytical stage.
• Development and provision of the lib5C software package for flexible analysis.

Conclusions:

• The choice of computational methods significantly influences 5C data analysis outcomes.
• The lib5C suite offers a valuable resource for optimizing 5C data interpretation.
• The principles derived from this analysis are applicable to other proximity ligation-based genomic interaction datasets like Hi-C, 4C, and Capture-C.