Bayesian Hierarchical Compositional Models for Analysing Longitudinal Abundance Data from Microbiome Studies

Figure 3

Model estimation and prediction of the female dataset at genus taxonomic level. Black points represent the data, the black line represents the expected value obtained with the proposed model, and dashed lines represent the value of . The vertical black line represents when the estimation of the model finishes and the prediction starts. The genera are, from left to right, o_bacteroidales;f_;g_, other, o_clostridiales;f_;g_, f_lachnospiraceae;g_, g_blautia, f_lachnospiraceae;g_clostridium, g_coprococcus, g_dorea, g_eubacterium, g_roseburia, g_ruminococcus, f_ruminococcaceae;g_, f_ruminococcaceae;g_clostridium, g_phascolarctobacterium, g_desulfovibrio, f_erysipelotrichaceae;g_clostridium, g_akkermansia, and g_bacteroides.