Description
Microbial composition varies considerably among different built environments, but the role of building materials in determining microbial diversity is not known. The overall goal of this study was to determine how building materials affect microbial growth and diversity under conditions of high humidity, and how bacterial and fungal community interplay. Our specific goal in this study was to determine how the incorporation of bacterial cell counts and fungal biomass estimates into sequence-based microbial community analysis affected the interpretation of microbial growth patterns on different building materials. To do this, we analyzed bacterial and fungal growth over time on 2 different building materials (gypsum drywall and medium density fiberboard) kept at high humidity. Deep sequencing of 16S and ITS libraries from DNA extracted from swabs was performed to determine bacterial and fungal composition, combined with bacterial cells counts and fungal spore and hyphae length data collected from an earlier study. With the sequence data, we performed taxonomy profiling, differential expression analysis and co-occurrence analysis to determine the impact of different building materials on microbial community. We observed that gypsum drywall is a more favorable environment not only for bacteria and fungi growth but also for the interplay between these two communities. We also identified a strong negative correlation between Bacillus and Geobacillus caused by Penicillium, a prevalent fungal genus, instead of assumed competition within bacterial community. This addresses the importance of global bacterial-fungal co-occurrence analysis for future built environment studies. To determine how quantitative data affect interpretation of microbial dynamics, we performed a quantitative profiling of bacterial and fungal community by multiplying count data to relative abundances. We observed that species abundance change is more significant when using quantitative profiling. We were also able to observe more true-positive and fewer false-negative correlations with quantitative profiling. Therefore, we conclude that the incorporation of quantitative profiling could provide novel insights into built environment dynamics.
