Strain-level typing of Wickerhamomyces anomalus using Fourier transform infrared spectroscopy and whole-genome sequencing

Abstract

In industrial settings, identifying the source of microbial contamination is crucial for effective microbiological risk assessment. While various strain identification technologies exist, many struggle with practicality, accuracy, and reproducibility. Fourier Transform Infrared Spectroscopy (FT-IR) has emerged as a rapid method, demonstrating a strong correlation with whole-genome sequencing (WGS) for certain bacteria. However, its accuracy for identifying yeast strains has been limited.

This study focuses on improving the accuracy of FT-IR for yeast strain identification by optimizing pretreatment conditions. We conducted phylogenetic analyses on Wickerhamomyces anomalus using both WGS single-nucleotide polymorphisms (SNPs) and FT-IR. Although initial FT-IR results were less accurate than WGS, refining the culture and sample preparation conditions led to significant improvements. We tested 16 different conditions, using Euclidean Distances (EDs) and dendrogram comparisons to evaluate discrimination ability, including metrics like the F-measure and adjusted Rand index (ARI).

The most accurate and reproducible FT-IR results were achieved with incubation in Sabouraud dextrose (SD) broth aligning closely with WGS results. This optimized FT-IR protocol now allows for rapid and precise strain-level discrimination of W. anomalus, offering a practical tool for tracking contamination sources in industrial environments.