Typically, modelling studies are performed with compounds from a single experiment (most of the time), synthesized and evaluated under a defined experimental condition. However, it's not always possible to have sufficient data from a single experiment for a specific target protein. I wonder if there is a way to mathematically convert the activity data of compounds from different experiments (with different experimental conditions) for QSAR studies.
There is nothing wrong about combining data from different experiments in a single structure-activity analysis study. The only prerequisite is to ensure that the data from different sources reflect the same biological effect (e.g., substrate-competitive enzyme inhibition). Indeed, sometimes, a mathematical formula can be used to convert one kind of readout into another one. For example, under certain conditions, Cheng-Prusoff equation can be used to calculate binding affinity from half maximal inhibitory concentration (IC50). Sometimes, activity data from different experiments may be inherently incompatible (e.g., allosteric and agonist-competitive effect on a receptor).
- Badges
- Science topic
- Similar topics
- Chemistry