Skip to main content
. 2021 Aug 10;10:e64644. doi: 10.7554/eLife.64644

Table 7. Random forest analysis.

One hundred random perturbations of all molecule concentrations (between +10% and −10%) were simulated. Features are ranked from most important to least important in predicting the temporal sensitivity. The sum of features’ importance equals 1.

Molecules abbreviation Weights Molecules full name
RasGRF 0.084 Ras guanine nucleotide–releasing factor
Sos 0.074 Son of sevenless
Cam 0.068 Calmodulin
MEK 0.050 Mitogen-activated protein kinase kinase
PDE2 0.050 Phosphodiesterase 2
pmca 0.048 Plasma membrane calcium pump
Epac 0.046 Exchange protein directly activated by cAMP
Calbin 0.046 Calbindin
CRKC3G 0.044 v-CRK Proto-Oncogene/Rap guanine nucleotide exchange factor 1
rasGap 0.043 Ras GTPase-activating protein
SynGap 0.042 Synaptic rasGap
PP1 0.038 Protein phosphatase 1
Shc 0.035 Src homology collagen-like
Grb2 0.034 Growth factor receptor-bound protein 2
PP2A 0.033 Protein phosphatase 2
MKP1 0.033 Mitogen-activated protein kinase phosphatase 1
Ng 0.033 Neurogranin
Src 0.032 Non-receptor tyrosine kinase
Cbl 0.030 Casitas B-lineage lymphoma
ERK 0.027 Mitogen-activated protein kinase
PDE4 0.025 Phosphodiesterase 4
PKA 0.025 Protein kinase A
NCX 0.022 Sodium calcium exchanger