Next-generation data filtering in the genomics era

Area of Science:

• Genomics
• Bioinformatics
• Population Genetics

Background:

• Genomic datasets are widely used but often contain errors or missing information, impacting analysis reliability.
• Data filtering is crucial for improving genomic data quality for downstream computational analyses.

Purpose of the Study:

• To review and suggest best practices for genomic data filtering.
• To improve implementation, reproducibility, and reporting standards for filter types and thresholds.
• To highlight the impact of filtering choices on population genetics statistics.

Main Methods:

• Review of common genomic data filters: minor allele frequency, missing data, linkage disequilibrium, and Hardy-Weinberg deviations.
• Illustration of filter effects using simulated and empirical datasets.
• Analysis of impacts on population genetics statistics like Tajima's D, FST, nucleotide diversity, and effective population size.

Main Results:

• Different filtering thresholds significantly affect population genetics statistics.
• Best practices are needed to standardize genomic data filtering.
• Careful selection of filters and thresholds is essential for reliable genomic analyses.

Conclusions:

• Standardized and reproducible genomic data filtering is essential for accurate downstream analyses.
• Researchers must carefully consider filter types and thresholds to ensure data quality.
• Adopting best practices will advance the next generation of genomic data analysis.