オーラルセッション
- 第2日 5月12日（火） 9:30～9:50 B会場（1202）
-
2B-O5-0930（2P-38） PDF
微生物由来揮発性有機化合物（MVOC）のイオン移動度およびGCMSデータの多変量解析による生育早期の土壌由来カビ種の同定II.
Microbial contamination often causes serious damage to cultural properties. It is imperative to develop an instrument which can detect the activation of fungi at an invisible stage before damage occurs. Software for fungal species identification based on multivariate statistical analysis of IMS and GC-MS spectral data of MVOC, named MVOC Finder, was created and released by T. Takeuchi and colleagues as open source software. MVOC finder is validated by the Receiver Operating Characteristic Curve (ROC). In this paper, we will show the advantages in detection of invisible fungi of MVOC finder, which applies partial least square regression (PLS), over principal component analysis (PCA).