Problem:
Virtual screening of large ligand libraries against a protein target is now a common component of any drug discovery program. Several docking programs such as Glide, GOLD, Autodock-vina, etc, exist for this purpose.
In another publication, the authors Perola, Walthers and Charifson conducted a study to benchmark the sampling capability of Glide for virtual screening. The ligand database used for docking had approximately 10,000 compounds, and was 'spiked' with 206 compounds known to have activity against each target protein used for docking. The docking results could therefore be evaluated for the ability to identify the known active compounds. i. [2 pt] The enrichment factor for Glide ranking of the top 3% of the compounds was reported to be 10.5. How many actives were found in the top 3% of Glide-ranked compounds? ii. [2 pts] Consider that you have the task of setting up an experimental high-through-put screen of your company's proprietary chemical library of 100,000 compounds against HIV protease and have been given a budget of approximately $22,500. The experimental cost is approximately $4500 for each 96-well plate setup to run the experimental screen. Based on your past experience, you assume there are approximately 500 compounds in the library with activity against HIV protease. Assuming the enrichment factor of 10.5 applies and is constant for the ranking of up to 10% of the library, how many hits (i.e. active compounds) do you expect to find within your budget?