Calibrating genomic and allelic coverage bias in single-cell sequencing

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Whole-genome amplification (WGA) is crucial for single-cell DNA analysis but introduces artifacts.
• These artifacts complicate accurate genomic information retrieval from single cells.
• Existing analytical strategies for bulk genomes are insufficient for single-cell data.

Purpose of the Study:

• To develop statistical methods for quantitatively assessing amplification bias in single-cell WGA.
• To benchmark different WGA technologies (MDA vs. MALBAC).
• To enable efficient and accurate variant detection from low-input samples.

Main Methods:

• Quantitative assessment of genome coverage bias at the amplicon level.
• Calibration of coverage bias magnitude using low-pass sequencing data.
• Benchmark comparison of multi-strand displacement amplification (MDA) and multiple annealing and looping-based amplification cycles (MALBAC).
• Development of statistical models for allelic bias calibration.
• Implementation of a census-based strategy for variant detection.

Main Results:

• Identified universal features of genome coverage bias in single-cell DNA libraries, predominantly at the 1-10 kb amplicon level.
• Demonstrated accurate calibration of coverage bias magnitude from low-pass sequencing (∼0.1×).
• Enabled prediction of depth-of-coverage yield for single-cell libraries at arbitrary sequencing depths.
• Provided a benchmark comparison of MDA and MALBAC single-cell libraries.
• Developed models to calibrate allelic bias and a strategy for accurate variant detection.

Conclusions:

• Statistical methods can accurately quantify and calibrate amplification bias in single-cell WGA.
• Low-pass sequencing is sufficient for predicting coverage yield.
• The developed models and strategies facilitate efficient and accurate variant detection from low-input samples, improving single-cell genomic analysis.