Application of OU processes to modelling temporal dynamics of the human microbiome, and calculating optimal sampling schemes

Table 6 Estimated standard deviations of \({\hat{\eta }}\) for various data sets and true mean reversion rates

Study	Body site	\(\eta =0.4\)	\(\eta =1\)	\(\eta =1.5\)	\(\eta =2\)
Moving picture	Person 1 Gut	0.093985	0.229590	0.404363	0.682644
	Person 1 Tongue	0.093338	0.224076	0.392306	0.660776
	Person 1 L. palm	0.093542	0.225305	0.395091	0.665951
	Person 1 R. palm	0.093542	0.225305	0.395091	0.665951
	Optimal sampling 131 samples	0.086844	0.217093	0.325639	0.434185
	Optimal equal-spaced 131 samples	0.087189	0.217972	0.326959	0.435945
	Person 2 Gut	0.058055	0.143934	0.256814	0.436208
	Person 2 Tongue	0.056130	0.133687	0.234737	0.395993
	Person 2 L. palm	0.056856	0.137175	0.242194	0.409543
	Person 2 R. palm	0.056524	0.135688	0.239088	0.403997
	Optimal sampling 351 samples	0.052923	0.132307	0.198460	0.264614
	Optimal equal-spaced 351 samples	0.053137	0.132843	0.199265	0.265686
David et al.	Subject A	0.060678	0.146476	0.258630	0.437380
	Optimal sampling 314 samples	0.055963	0.139909	0.209863	0.279817
	Optimal equal-spaced 314 samples	0.056190	0.140476	0.210713	0.280951
	Subject B	0.078551	0.195710	0.347970	0.589763
	Optimal sampling 183 samples	0.073391	0.183477	0.275215	0.366953
	Optimal equal-spaced 183 samples	0.073688	0.184220	0.276330	0.368441

ISSN: 1471-2105