Additional file 2: of Measuring associations between the microbiota and repeated measures of continuous clinical variables using a lasso-penalized generalized linear mixed model
Laura Tipton
Karen Cuenco
Laurence Huang
Ruth Greenblatt
Eric Kleerup
Frank Sciurba
Steven Duncan
Michael Donahoe
Alison Morris
Elodie Ghedin
10.6084/m9.figshare.6553952.v1
https://springernature.figshare.com/articles/journal_contribution/Additional_file_2_of_Measuring_associations_between_the_microbiota_and_repeated_measures_of_continuous_clinical_variables_using_a_lasso-penalized_generalized_linear_mixed_model/6553952
Figure S1. Additional OC-COPD associations between laboratory measurements and bacteria identified by LassoGLMM. Strong associations between bacteria and (a) percent neutrophils (O1), (b) partial pressure of oxygen PO2(O4) (c) SAT (O5), (d) alkaline phosphatase (O6), and (e) glucose (O10). Each horizontal grey line represents an individual. When a colored circle is located on the grey line, it is the relative abundance of that microbe for that subject. Perfect positive association between clinical variable and bacteria would form a line from the bottom-left to the top-right of the graph and would have a highly positive β coefficient in the LassoGLMM. Perfect negative association would form a line from the top-left to the bottom-right of the graph and would have a highly negative β coefficient. (PDF 287 kb)
2018-06-15 05:00:00
Microbiota
16S
ITS
Repeated measures
Continuous outcomes
GLMM
Lasso