Table 2. .

Relation between the effect of virtual resections of the RA (as given by the normalized decrease in spreading after the resection) and the properties of the model, baseline, and postresection networks

Baseline	Effect	r 2	p	Resection	Effect	r 2	p
βκ/γ	+	0.09	0.01	E seed	−	0.10	0.006
S RA	+	0.33	8 · 10−8	ΔEseed	+	0.06	0.04
S seed	−	< 10−2	0.7	BCseed	−	0.11	0.004
E RA	+	0.11	0.004	ΔBCseed	+	0.15	5 · 10−4
E seed	−	0.02	0.2	c seed	−	0.21	3 · 10−5
BCRA	+	0.08	0.01	c Netw	−	< 10−2	0.5
BCseed	−	0.05	0.06	Δcseed	+	0.27	2 · 10−5
c RA	+	0.003	0.6	ΔcNetw	+	0.08	0.01
c seed	−	0.11	0.004	F seed	−	0.21	3 · 10−5
c Netw	−	0.002	0.7	F Netw	−	< 10−2	0.8
F RA	+	< 10−2	0.8	ΔFseed	+	0.29	7 · 10−7
F seed	−	0.02	0.2	ΔFNetw	+	0.11	0.003
F Netw	−	2 · 10−5	0.9

Note. As model parameter we consider the spreading-to-recovery ratio βκ/γ, which combines the three model parameters in one. The preresection network was characterized by the size S, the out-connectivity E, the betweenness centrality BC, the clustering c, and the efficiency F of the RA and the seed, and by the average clustering coefficient and efficiency of the network. The resected network (i.e., the network after the virtual resection of the RA was performed) was characterized by the new out-connectivity, BC, clustering, and efficiency of the seed, by the network-averaged clustering and efficiency, and by their respective decreases due to the virtual resection, that is, ΔX = X(baseline) − X(resection). Significant effects (p > 0.05) are indicated by bold font in the p value.