Skip to main content
. 2019 Mar 18;9:4700. doi: 10.1038/s41598-019-41072-w

Table 2.

Prediction models for selected rheological attributes.

Rheological Attribute Prediction Formula PLS factors VIP R²Y (%)
Log10(Jmax) 1.598pw48.694br+0.016apl9.188ly+52.705epr4.874 4 apl
ly
89.65
log10(Jr) 0.950pw29.975br+0.005apl6.651ly+3.072epr2.905 3 apl
ly
81.56
J el 33.843pw+689.778br0.265apl+72.428ly1591.011epr+132.938 4 pw
br
apl
ly
epr
71.85
Log10(η0) 57.529br1.567pw0.018apl+11.776ly54.449epr+6.519 4 apl
ly
89.68
1/G* 9.275105pw2.446103br+8.503107apl4.547104ly+3.391103epr8.588105 5 br
ly
84.26
tanδ 0.151pw4.067br+7.868104apl1.010ly+1.501epr+0.377 3 aplly 84.83
1/Af 9.161105pw2.491103br+8.597107apl4.586104ly+3.489103epr8.525105 5 pw
br
ly
83.81
z 2.222pw+81.507br+5.997103apl+13.389ly29.568epr5.274 4 apl
ly
83.26

Equations for rheology prediction were determined by PLS models of the microstructural attributes protein width (pw), branching rate (br), average protein length (apl), lacunarity (ly) and end-point rate (epr). The models’ dominated attributes are given by VIP (variable importance for projection) values greater than 0.8. R²Y describes the percentage explained for cumulative Y.