Skip to main content
. 2019 Apr 30;9:6711. doi: 10.1038/s41598-019-43160-3

Table 3.

Model selection table is shown for the linear models of bank vole infection status (infection).

Model Model structure df logLik AICc Delta Weight (%)
model011 infection~E + N + E:N 5 −12.287 35.212 0.000 44.374
model007 infection~E + S + N + E:N 6 −11.802 36.507 1.295 23.226
model005 infection~E + S + N + E:N + S:N 7 −11.164 37.546 2.334 13.817
model003 infection~E + S + N + E:S + E:N 7 −11.317 37.851 2.639 11.862
model002 infection ~E + S + N + E:S + E:N + S:N 8 −11.032 39.646 4.434 4.835
model001 infection~E + S + N + E:S + E:N + S:N + E:S:N 9 −10.826 41.651 6.439 1.774
model014 infection~E + N 4 −21.428 51.277 16.065 0.014
model606 infection~geno6 + S + N + geno6:S 6 −19.780 52.463 17.251 0.008
model009 infection ~E + X + N 5 −20.939 52.517 17.305 0.008
model008 infection ~E + X + N + S:N 6 −19.871 52.645 17.433 0.007

The analysis was done on the entire data set of 100 bank voles. Of the 117 models, the top 10 models are shown; the top 6 models have 99.9% of the support. Table S11 in the ESM shows all 117 models. The fixed factors are number of engorged B. afzelii-infected nymphs (N), experiment (E), sex (S), and TLR2 genotype. TLR2 genotype is modelled in 7 different ways (geno1, geno2, geno3, geno4, geno5, geno6, and geno7). For each model, the model structure, model degrees of freedom (df), log likelihood, corrected AIC value (AICc), difference in AICc from the top model (Delta), and the support (Weight) expressed as a percent are shown.