Figure S8. Increasing the time series length improves the accuracy of the test for temporal structure. Noise types were computed for time series sub-sets from 1000 to 1050 (a) and 1000 to 1100 (b) for all data sets with more than 1000 time points. Labels for time series are colored according to the level of non-zero intrinsic noise (sigma) for Ricker, according to the death rate if larger than one for Hubbell, according to the interval if larger than one (with interval coloring taking precedence over sigma) and black otherwise. (PDF 7 kb
Irregular sub-daily oscillation observed in PhILR balance n12 does not correlate with known external...
NMDS plots using the Bray distance metric with all the samples except for the Yatsunenko study sampl...
Control analysis of Mystic Lake and control samples. a.) PCoA plot of samples colored according to d...
Figure S7. Increasing the time series length improves the accuracy of the test for temporal structur...
Figure S5. The test for temporal structure with noise types is robust to compositionality and the ab...
Figure S6. The test for temporal structure with noise types is robust to noise. (a) Noise-type profi...
Figure S4. The noise-type classification and the neutrality test are robust for a wide parameter ran...
Figure S1. Maximal autocorrelation and Hurst exponent profiles reproduce patterns seen with noise ty...
Figure S2. The noise-type classification and the neutrality test for Ricker and gLV are robust to po...
Table S1. lists for each test time series the parameters used to generate it (sheet “Model parameter...
Figure S11. Time series of the 100 top abundant OTUs in the processed stool data of individual A and...
Figure S10. The accuracy of network inference with LIMITS decreases more strongly when applied to th...
Figure S3. The noise-type classification and the neutrality test for SOI are robust to interaction m...
Figure S12. Variability of noise-type classification across rarefactions. The noise types of 100 tax...
BACKGROUND: Growth rates, interactions between community members, stochasticity, and immigration are...
Irregular sub-daily oscillation observed in PhILR balance n12 does not correlate with known external...
NMDS plots using the Bray distance metric with all the samples except for the Yatsunenko study sampl...
Control analysis of Mystic Lake and control samples. a.) PCoA plot of samples colored according to d...
Figure S7. Increasing the time series length improves the accuracy of the test for temporal structur...
Figure S5. The test for temporal structure with noise types is robust to compositionality and the ab...
Figure S6. The test for temporal structure with noise types is robust to noise. (a) Noise-type profi...
Figure S4. The noise-type classification and the neutrality test are robust for a wide parameter ran...
Figure S1. Maximal autocorrelation and Hurst exponent profiles reproduce patterns seen with noise ty...
Figure S2. The noise-type classification and the neutrality test for Ricker and gLV are robust to po...
Table S1. lists for each test time series the parameters used to generate it (sheet “Model parameter...
Figure S11. Time series of the 100 top abundant OTUs in the processed stool data of individual A and...
Figure S10. The accuracy of network inference with LIMITS decreases more strongly when applied to th...
Figure S3. The noise-type classification and the neutrality test for SOI are robust to interaction m...
Figure S12. Variability of noise-type classification across rarefactions. The noise types of 100 tax...
BACKGROUND: Growth rates, interactions between community members, stochasticity, and immigration are...
Irregular sub-daily oscillation observed in PhILR balance n12 does not correlate with known external...
NMDS plots using the Bray distance metric with all the samples except for the Yatsunenko study sampl...
Control analysis of Mystic Lake and control samples. a.) PCoA plot of samples colored according to d...