Is it necessary to calculate lane normalisation factor when doing western blot data analysis? I have recently done my first western blot and I am doing data analysis to quantify my blot. I have labelled my membrane against inactive GSK3 and active GSK3 which are phosphoproteins so I am using total GSK3 as an internal loading control. I have read some guides and handbooks about western blot data analysis and I have seen that some of them calculate a lane normalisation factor to account for variations in signal intensities of the loading control. For all loading control bands in each lane, they divide by the loading control band with highest intensity to get the lane normalisation factor. Then for each band of the target protein of interest they divide by the lane normalisation factor to get the normalised intensity. I was wondering in general with western blots is it good practice to calculate the lane normalisation factor when doing the data analysis? Any insights are appreciated.

Generally speaking, the proper way to quantify a western blot is to normalize to a loading control such as Actin or GAPDH. In this case it would be (pGSK3/Actin)/(GSK3/Actin) as total GSK3 is not a loading control. A loading control is to protein that is accepted as unvarying in concentration across multiple samples if the same protein amoun is used, and to my knowledge GSK3 would show more biological variability than accepted loading controls. This would be considered the best practice to compare protein concentrations across samples.

That being said, if all you care about is the ratio of active to total, pGSK3/Total GSK3 is probably sufficient to assess if your manipulation changed active/inactive:total ratios. For example, in autophagy assays you can measure autophagic flux by measuring LC3-I to LC3-II, and I've never seen them normalized to a loading control since its a ratiometric readout contained within a single sample.