Regression model

Dichotomous assay documents currently support a generalized linear model for parallel-line potency assays using quantal (count) data, with a probit or logit link function.

Probit

Uses the Probit regression model to estimate relative potency.

The model is defined by the following equation:

$$\text{P(}\mathrm{\text{Dose}}\text{)}=\text{Φ}\left(\text{m}\cdot {\mathrm{\text{log}}}_{\text{b}}\text{(}\mathrm{\text{Dose}}\text{)}+\text{n}\right)$$

Logit

Uses the Logit regression model to estimate relative potency.

The model is defined by the following equation:

$${\mathrm{\text{log}}}_{\text{b}}\left(\frac{\text{P(}\mathrm{\text{Dose}}\text{)}}{1-\text{P(}\mathrm{\text{Dose}}\text{)}}\right)=\text{m}\cdot {\mathrm{\text{log}}}_{\text{b}}\text{(}\mathrm{\text{Dose}}\text{)}+\text{n}$$