| Autor: |
Hegarty B; Department of Chemical and Environmental Engineering, Yale University, P.O. Box 208263 New Haven, Connecticut 06520-8286, United States., Pan A; Department of Chemical and Environmental Engineering, Yale University, P.O. Box 208263 New Haven, Connecticut 06520-8286, United States., Haverinen-Shaughnessy U; Indoor Air Program, The University of Tulsa, 800 South Tucker Drive, Henneke 212, Tulsa, Oklahoma 74101-9700, United States., Shaughnessy R; Indoor Air Program, The University of Tulsa, 800 South Tucker Drive, Henneke 212, Tulsa, Oklahoma 74101-9700, United States., Peccia J; Department of Chemical and Environmental Engineering, Yale University, P.O. Box 208263 New Haven, Connecticut 06520-8286, United States. |
| Abstrakt: |
Dampness or water damage in buildings and human exposure to the resultant mold growth is an ever-present public health concern. This study provides quantitative evidence that the airborne fungal ecology of homes with known mold growth ("moldy") differs from the normal airborne fungal ecology of homes with no history of dampness, water damage, or visible mold ("no mold"). Settled dust from indoor air and outdoor air and direct samples from building materials with mold growth were examined in homes from 11 cities across dry, temperate, and continental climate regions within the United States. Community analysis based on the sequence of the internal transcribed spacer region of fungal ribosomal RNA encoding genes demonstrated consistent and quantifiable differences between the fungal ecology of settled dust in homes with inspector-verified water damage and visible mold versus the settled dust of homes with no history of dampness, water damage, or visible mold. These differences include lower community richness ( p adj = 0.01) in the settled dust of moldy homes versus no mold homes, as well as distinct community taxonomic structures between moldy and no mold homes (ANOSIM, R = 0.15, p = 0.001). We identified 11 Ascomycota taxa that were more highly enriched in moldy homes and 14 taxa from Ascomycota , Basidiomycota, and Zygomycota that were more highly enriched in no mold homes. The indoor air differences between moldy versus no mold homes were significant for all three climate regions considered. These distinct but complex differences between settled dust samples from moldy and no homes were used to train a machine learning-based model to classify the mold status of a home. The model was able to accurately classify 100% of moldy homes and 90% of no mold homes. The integration of DNA-based fungal ecology with advanced computational approaches can be used to accurately classify the presence of mold growth in homes, assist with inspection and remediation decisions, and potentially lead to reduced exposure to hazardous microbes indoors. |
