molecular docking algorithms provide readable image about how can the antagonist-receptor and predict bio activity of your antagonist.

Hi,

The agonist or antagonist effects are linked to conformational changes in the receptor, so more than docking, the importance would consist in being able to correlate theoretical and experimental results (structural data from the PDB) . The theoretical part could be explored through the evaluations of the affinity and conformational changes upon ligand recognition of agonist and antagonist ligands through the combination of docking and molecular dynamics studies.

The order should not matter, you can do both. For example you can use molecular docking to prioritize the compounds that should be tested. Or you can conduct the in vivo experiment and then use molecular modelling to elucidate more information such as binding mode/site, interacting residues, SAR, etc. Just as previously stated, it is desirable that both theory and experiment point towards the same results, beware that this is not always the case though. Be extra careful when working with chiral molecules as these often give unexpected results/artifacts.

It is relevant if you know the key residues and type of interactions relates to the active/inactive state of the targeted receptor. For some receptors, as some in the family A of GPCR key interactions have been mapped by multiple experimental and theoretical approaches.

It is a prediction study for the activity mechanism. Also it has more valuable when correlate with in vitro or in vivo study with considering the other factors effect.