A big part of the interpretation depends on: what question did you ask?  Did you dock a single molecule?  Did you dock a library?  Were you doing drug screening?  Were you trying to determine how good the algorithm would be for your use case?

Assuming a typical high-throughput screen: start by looking at the docking scores.  These are not as quantitatively ranked as one would imagine, but they're still a fine place to start.  Strip the scores out of the files you generated with 'grep', or some similar utility, load them into Excel (or something else), and sort by score.  This will tell you which molecules were predicted to bind best.  Again - this rank order won't actually match what a biological assay would give you, but it's a good starting point.

From there, look at the best molecules, make sure they actually look right.  By this, I mean make sure that amide bonds are trans, benzene rings are planar, etc.  Make sure there's good shape complimentarity to the protein.  If you see multiple similar poses of the same molecule highly ranked, that's a good sign.

Look for hydrogen bonds.  See how ideal they are.  If you have to decide between poses/molecules with more or less bonds, I tend to trust the ones that make more hydrogen bonds to be a more accurate reflection of the likely binding pose than those that depend more upon hydrophobic interactions.

From there, run assays on the most promising looking molecules to see which ones actually do anything.  But again: this is basic advice for one common use scenario.  What you want to do may differ if you asked a different question.

Thank you William P Katt. Your questions and response is really helpful. Concerning what I docked, I docked a Library and yes I am actually screening to find out potential HIV-1 Protease Inhibitors Inhibitors from the Library(FDA Approved drugs).

I already have the scores stripped and sorted the scores with Excel and with that, I have the molecules predicted to bind best.

What I don't know right now is what to do next and how to carry out the analysis of the results.

Your analysis is basically done. Now, you may consider two following options: 1) discuss the results with a medicinal chemist (the one who is supposed to follow up on the virtual screening hits). He might express certain preferences with respect to your top ranking hits; 2) Use more accurate affinity predictors (e.g., MM-PBSA) to re-rank the top scorers.

After 1 and/or 2 (or skipping both), you can procure and experimentally screen the top scorers.