The estimation of within-subject covariate effects is robust against between-subject confounders, and longitudinally measured microbiome data enable the identification of microbiota effects on the risk of diseases in the host. As most existing studies are cross-sectional in nature, the validity and interpretation of their results are limited. Therefore, longitudinal studies are needed to investigate the association between the human microbiome and diseases. However, high inter-subject variation and compositionality induces type-1 error inflation and it make the use of traditional methods for longitudinal analysis hard. Type-1 error inflation is deepen without considering the correlation within subjects. The proposed mTMAT provides test statistics for each node, which are combined to identify mutations associated with host diseases and considers correlations among the repeately observed measures.
Authors
Kangjin Kim, Ph.D <rekki@channing.harvard.edu>
Sungho Won, Ph.D <won1@snu.ac.kr>
Written and maintained by Kangjin Kim.
| mTMAT.R | R function for mTMAT |
| User Reference Guide | A description for R function and example code |
| Example metagenome data file | An example metagenome data |
| Example metadata file | An example metadata |
| Example taxonomy file | An example file contatining taxonomy information |
| Example phylogentic tree file | An example file of phylogenetic tree |