Bio-Rad Image Lab™ Software - Manual (Page 204)

E | Regression Calculation Methods
204 | GS-900 Calibrated Densitometer with Image Lab Software
where:
x = mobility
y = molecular weight
a = estimated molecular weight at infinity
b = slope of the tangent at midpoint
c = midpoint
d = estimated molecular weight at zero mobility
Since the curve generated by the logistic-4PL regression method represents a
perfectly shaped S, it might not fit the data very well in all cases.
Cubic spline: Cubic spline curves are smooth curves that go through every data
point. The model is a cubic polynomial on each interval between data points. In
some cases, a spline curve can work well as a standard curve for interpolation.
However, because the curve is calculated individually for every pair of points, it does
not correspond to any single equation.