ntsm: an alignment-free, ultra-low-coverage, sequencing technology agnostic, intraspecies sample comparison tool for...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Sample swapping is a common error in large cohort studies, particularly with heterogeneous data types from various sequencing technologies.
• Current sample swap detection methods are computationally expensive, requiring data alignment, sorting, and indexing, which is often unnecessary for downstream analyses like genome assembly.
• The increasing scale and complexity of genomic studies necessitate robust and efficient quality control tools.

Purpose of the Study:

• To develop a faster and more resource-efficient method for detecting sample swaps in large-scale genomic studies.
• To provide a quality control tool applicable to diverse sequencing data types without requiring data alignment.
• To offer additional QC metrics such as error rates and coverage bias.

Main Methods:

• Utilizes indexed k-mer sequence variants to determine sample similarity.
• Incorporates coverage information at variant sites to enhance statistical power via a likelihood ratio-based test.
• Employs spatially indexed principal component analysis (PCA)-based prescreening to accelerate analysis by avoiding exhaustive comparisons.

Main Results:

• The method accurately identifies sample swaps by analyzing k-mer variants and coverage data.
• It enables estimation of per-sample error rates and coverage bias.
• PCA-based prescreening significantly speeds up the analysis process.

Conclusions:

• This novel tool processes raw sequencing data, bypassing the need for alignment, thus saving significant computational resources in standard quality control (QC) pipelines.
• The method is robust across different sequencing data types (e.g., Oxford Nanopore, Pacific Biosciences, Illumina), making it suitable for studies combining multiple technologies.
• Beyond sample swap detection, it provides valuable QC information and facilitates population-level PCA for ancestry analysis visualization.