In mixed models, odor increased 0.15 ± 0.05 units (mean ± SE) for a 1-ppb increase in H2S, and 0.45 ± 0.14 units for a 10-μg/m increase in PM10 at wind speeds > 6.75 miles per hour.
Including focal plant identity as a random factor in the general linear mixed models substantially increased model fit for 8 of the 10 response variables [see Supporting Information—Table S1 ], explaining between 20 and 80%% of the variation in a given response variable [see Supporting Information—Table S1 ].
The same topology was found in both MP and Bayesian analyses irrespective of utilizing codon positions for the Bayesian cytB analyses and also for each of the two multi-locus datasets; however, the mixed models provided increased support indices at most nodes for all data sets, and therefore, only the support indices while utilizing codon partitions are shown for the Bayesian results (Figs. 2, 3).
Compared with a simple mixed linear model, compressed mixed linear models increase the statistical power and reduce computing time for large samples [ 74].
Our analytical strategy for this large data set centered on using a series of mixed models of increasing complexity to elucidate the extent of pleiotropy among gene expression traits.
This difference is expected to further increase when the number of records used in the analysis increases, because the size of the G matrix and therefore the size of the left-hand sides of the mixed model equations increases quadratic with the number of animals, while the number of calculations in the Bayesian models increases less than linearly.
Such mixed model analyses allow increased statistical power in longitudinal dataset such as ours, with variable time between visits and a large age range at T1 evaluation.
"Efficient Bayesian mixed model analysis increases association power in large cohorts". Nature genetics 47 (3): 2847290.
According to our mixed model estimates, MtC increases were statistically lower than the rise in mortality.
Our study illustrates the utility and advantages of using more highly time-resolved exposure and outcome data, which not only increases statistical power in mixed models because of an increased sample size but also captures the often significant variability in pollutant concentrations over the course of the day.
