-Once the ligand binding has been quantified, the total amount of ligand bound does not necessarily mean that all the ligand has bound to the receptor, and the amount bound will almost certainly include an element of non-specifically bound ligand. The contribution of non-specific binding to the total is usually estimated by inclusion of control experiments in which a huge excess of unlabelled ligand, approximately a 100-fold excess, is added. This means that the high affinity receptor sites will be saturated with unlabelled ligand, and, therefore, any labelled ligand remaining bound will be a result of binding to non-specific sites. The non-specific binding is, in general, linear in a concentration-dependent manner. Once these values are subtracted from the total binding curve, the binding specific to the receptor can be seen. -This binding starts off being very much greater than the non-specific binding, because of the high affinity of the receptors, but, like typical enzyme kinetics, the receptors become saturated and the binding curve tails off to a maximum, beyond which no more ligand binding can occur despite the addition of more ligand. Therefore, from this graph, the total specific binding sites can be estimated, as can the Kd value. -Estimating the binding characteristics from a curve can be difficult and not very meaningful, and, therefore, in an analogous way to the analysis of enzyme activity, the ligand binding data must be mathematically manipulated to give a linear relationship. Such analysis should give a better insight into the characteristics of the binding, for example, whether the binding shows any cooperativity, that is, whether the binding of the second ligand is more or less favourable because of the bound first ligand. The two common methods employed are derived from those developed by Hill and Scatchard . The original development of the Hill plot was to analyse the binding of oxygen to haemoglobin , but the same rationale can be used here.