- 第3日 6月24日（金） 9:40～10:00 C会場（411，412）
Multivariate analysis such as principal component analysis (PCA) and partial least squares (PLS) has been widely applied in metabolomics. After visualizing metabolome data, important metabolites for biological interpretation are selected using principal component loadings and PLS loadings. However, there are few tools to calculate these loadings, and these loadings are still not fully utilized. So, we released R package “loadings” in 2021 to select significant metabolites by using PC and PLS loadings. And PLS rank order of groups (PLS-ROG) and orthogonal smoothed PCA (OS-PCA) can be performed using “loadings” package. Currently, we have also implemented multiset PLS and multiset PLS-ROG to integrate multiple dataset such as multi-omics data or metabolome data derived from biological fluids and multiple organs. We introduce the results of the integration of COVID-19 proteome and metabolome data, and metabolome data of plasma, liver, cardiac muscle and brain in hyperlipidemic rabbits. Recently, large-scale metabolome data have been accumulated in public repositories, and it is expected that new biological findings will be obtained by integrating metabolome data. We are developing a new network-based approach to integrate large-scale metabolome data acquired on different platforms and at different research institutions, and we would like to introduce the results obtained to date.