Fast barcode calling based on k-mer distances.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• DNA barcodes are essential for identifying samples in pooled sequencing.
• Accurate barcode calling is critical but computationally intensive, especially with high error rates.
• Existing methods struggle with the scale of millions of barcodes and billions of reads.

Purpose of the Study:

• To develop an efficient and precise computational approach for DNA barcode calling.
• To address the challenges of speed and accuracy in large-scale sequencing data analysis.

Main Methods:

• A novel barcode calling approach utilizing a filtering technique based on precomputed k-mer lists.
• Implementation on server GPUs and CPU-parallel processing for high throughput.

Main Results:

• The proposed method achieves slightly higher accuracy than state-of-the-art approaches.
• It is over 500 times faster than existing methods.
• Enables calling for one million barcodes and one billion reads per day on a server GPU or CPU.

Conclusions:

• The k-mer filtering approach offers a significant advancement in barcode calling efficiency and accuracy.
• This method is highly scalable for massive sequencing datasets.
• Provides a practical solution for high-throughput pooled sequencing analysis.