To keep this question short, let's say I have a novel membrane protein with the known structural fold of a known transporter superfamily so I know its function is a transporter of cellular metabolites (Coenzyme-A, PLP) etc.
Also, say I have computationally docked Coenzyme-A and PLP to the binding site and the poses all have similar negative predicted binding affinity energies etc.
How would you experimentally test or validate that one of these molecules may be transported by the protein?
I am from a bioinformatics background with absolutely no biochemistry lab experience or knowledge, so any insights would be most appreciated, thanks.