Study of interaction in dual-species biofilm of Candida glabrata and Klebsiella pneumoniae co-isolated from peripheral venous catheter using Raman characterization mapping and machine learning algorithms

Polymicrobial biofilm infections, especially associated with medical devices such as peripheral venous catheters, are challenging in clinical settings for treatment and management. In this study, we examined the mixed biofilm formed by Candida glabrata and Klebsiella pneumoniae, which were co-isolated from the same peripheral venous catheter. Our results revealed that C. glabrata can form mixed biofilms with K. pneumoniae in vitro on peripheral venous catheters and the bottom of microplate wells, as confirmed by scanning electron microscopy. Additionally, using Raman mapping, we showed the distribution of both species in mono- and dual-species biofilms and suggested the type of microbial interaction in this polymicrobial biofilm. Finally, with the assistance of appropriate machine learning (ML) algorithms, based on identified peaks of bacteria, yeast, catheter, and Microplate mapping spectra, we develop a dedicated Raman database to detect the presence of these elements in an unknown spectrum in the future. •Candida glabrata formed mixed biofilms with Klebsiella pneumoniaein vitro, as confirmed by scanning electron microscopy.•The Raman microscopy mapping showed the distribution of both species in single and mixed biofilms.•Using AI tools, we developed a potential method to identify and quantify reference elements in unknown Raman spectra.