Error in Low-Coverage Genome Assemblies
Author Information
Author(s): Hubisz Melissa J., Lin Michael F., Kellis Manolis, Siepel Adam
Primary Institution: Cornell University
Hypothesis
What is the extent of sequencing error in low-coverage genome assemblies and its impact on downstream analyses?
Conclusion
Sequencing errors in low-coverage genome assemblies can significantly affect downstream analyses, but automatic error mitigation methods can help reduce their impact.
Supporting Evidence
- Sequencing error rates were estimated to be 1-4 errors per kilobase.
- Most errors were found in low-quality bases, particularly at the ends of reads.
- Automatic methods for error mitigation can mask or eliminate substantial fractions of errors.
Takeaway
When scientists look at genomes with low coverage, they often find mistakes. This study shows how to fix some of those mistakes to get better results.
Methodology
The study compared low-coverage genome assemblies with high-quality sequences from the ENCODE project to estimate sequencing error rates and explored automatic methods for error mitigation.
Potential Biases
Errors are concentrated in regions of single-read coverage, which may lead to biased interpretations in phylogenomic analyses.
Limitations
The study primarily focuses on low-coverage assemblies, which may not represent the full spectrum of sequencing errors in higher coverage assemblies.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website