I got two .pdb files for antigen and antibody. I need to dock them, but I don't have a file with antibody-antigen conjugate. Is it possible to compute binding?
it should be easily done using HDOCK http://hdock.phys.hust.edu.cn/ (assuming that both antigen and antibody are proteins). You will get 10 models sorted from best to worse. Let me know if it works for you
As Martin says, you can use the free online server above. However, if you want a bit more control and accuracy, then you might want to use something to probe the protein for likely binding sites first using something such as DeepSite (https://www.playmolecule.com/). This will identify likely binding sites which you can then focus your search on to produce more accurate binding predictions. You can perform slightly more detailed predictions using something such as Autodock Vina. If you can get hold of Schrodinger's Glide, then I recommend that, but it can be expensive.
I would like to add a bit to Stefan Bidula Sir, you can also use DogsiteScorer (https://proteins.plus/) to find out your active sites based on best drug score, rather than going for a blind interaction of your antigen and antibody.