A fast machine-learning-guided primer design pipeline for selective whole genome amplification

Area of Science:

• Microbial genomics
• Evolutionary biology
• Infectious disease research

Background:

• Analyzing microbial genome populations is crucial for understanding microbial evolution and pathogenesis.
• Current limitations in obtaining pure microbial DNA hinder next-generation sequencing for population genomics.
• Selective whole genome amplification (SWGA) is a key technique for obtaining sufficient DNA.

Purpose of the Study:

• To present swga2.0, an optimized and parallelized pipeline for designing selective whole genome amplification (SWGA) primer sets.
• To improve the efficiency and accuracy of primer set design for microbial population genomics.
• To overcome practical limitations in preparing microbial DNA for sequencing.

Main Methods:

• Developed swga2.0 pipeline incorporating active and machine learning for primer efficacy evaluation.
• Implemented optimized primer set search and evaluation strategies with parallelization.
• Used empirical data to identify primer characteristics that enhance amplification performance.

Main Results:

• swga2.0 significantly decreases pipeline runtime through parallelization.
• The pipeline successfully designs primer sets for effective SWGA.
• Demonstrated successful amplification of Prevotella melaninogenica DNA from human-dominated samples.

Conclusions:

• swga2.0 enhances the feasibility of microbial population genomic studies.
• The pipeline facilitates the analysis of microbial communities, including those in complex environments like cystic fibrosis lung microbiome.
• Improved SWGA primer design accelerates research in microbial evolution and pathogenesis.