This case study shows how to use MixKernel to explore environmental and metagenomic patterns in Tara Ocean data through Kernel PCA. MixKernel combines multiple kernel matrices derived from environmental data, OTU abundances, and gene functions, enabling the integration of diverse data types for unsupervised dimension reduction. You’ll learn how to compute and combine kernels, perform KPCA, assess variable importance, and visualise relationships between key microbial and environmental features.
🔍 More on MixKernel
📄 Download R script
References:
1. Mariette, J. and Villa-Vialaneix, N. (2017) Integrating TARA Oceans datasets using unsupervised multiple kernel learning. Bioinformatics 34(6)
2. Crone L. and Crosby D. (1995). Statistical applications of a metric on subspaces to satellite meteorology. Technometrics, 37(3), 324-328.