Teaching Points for Virtual Screening 3

VDS Staff,

These final two labs will be done as a pair. They can do one before the other if they want to.

We don’t know how smoothly the Virtual Screening will be since GOLD is acting up on the DDFE and isn’t using the full blades.

The link between the two lab is that some of the compounds from the virtual screening – we will be testing in the wet lab.

There are 2 GOLD runs to the lab

1st Run – 5,000 compounds

2nd Run – 500 compounds

This will be their first time using the GOLD GUI – i.e. the Hermes interface. So, it is their first time with Xming as well.

Make sure they aren’t all on the same blades (slower). So, we will help them select which blade to use based on the $ganglia cpu_user command.

Have them choose an empty blade on the upper deck (#8-15)

Using only 4 processors per run.


One big problem is that we have to use GOLD 5.0 for the Hermes and then switch back to GOLD 4.1 for the docking jobs. So, they need to use:



The blades are super slow. We may have to kill their jobs and use an alternative way to run GOLD (gold_scan that runs on the SGE system)

If their jobs get really messed up, we will need to help them kill them and then wipe the blades clean.

There is the long process that is listed in a file in the DatabasesVDS folder called ClusterCleaning.txt

Then there is the faster/easier way which is to run the script file: script_ScanKillUserPVMnum.sh

To run this file, login as the student, get into the directory DatabasesVDS/LabVS3_PTP1bLibrary, and type:

$sh name_of_script_file.sh

It will take a few minutes to run – but the blades for that student should be cleaned.


In drug discovery, we try to find novel compounds that have different structures and properties from drugs that are currently out there.

Different crystal structures and how they are prepared (hydrogens, charges, etc) can affect the docking results. That is why it is good to test out a few different ones.



Understanding file structure in Linux operating environment and how to move around in the directories

Looking up molecular properties of compounds and determining if they might make a good drug.


2 Responses

  1. Hey Dr. B,

    Will they need to take a picture of each of their 50 ligands for their data analysis? It seems a little much, so I wasn’t sure what to tell those that asked. I will email everyone once you let me know.


  2. Leighanna,

    They just need 3 PyMol pictures.
    They are listed inthe protocol as the Top Chembridge compound, The Top BindersDB compound, and then the one they tested in wetlab.

    Thanks, Dr. B

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: