DOCK Blaster:Progress Bar
The DOCK Blaster Progress Bar tells you where you are in the docking pipeline by coloring your current stage in yellow.
There are six stages, some with several substages, as follows: Preparer, Scrutinizer, Target Preparation, Calibration, Docking, and Results.
The preparer assists you in preparing a job. Here you upload your data to the DOCK Blaster server. For more information about the data formats, please see Prepare Input.
This is the page you see immediately after pressing the "DOCK" button. Your data are uploaded to the server, and the Scrutinizer checks the data for consistency. For more information about the Scrutinizer, and the many ways it can complain about your data, please see Scrutinizer.
This stage prepares "hot spots" (aka "orientation spheres") and precalculates scoring grids. It also prepares libraries of compounds for calibration calculations. There are three sub-stages of target preparation: Queued, Running, and Done.
Your job has been submitted to the queue, but has not yet started. It will start as soon as system capacity allows. If the cluster is heavily loaded, you can potentially be stuck in this stage for a long time. We aim to give you useful guidance to help you manage your expectations.
Your job is running, or at least, it has not reported as having ended. This stage typically takes 15 minutes, but can take an hour. If it takes much more than an hour then something has probably gone wrong.
Target preparation has ended normally. Normally, the calculation proceeds directly into calibration, so it is rare to be at this stage.
Calibration is the process of performing control calculations to investigate whether docking can recapitulate known experimental information. DOCK Blaster tries two different sampling schemes and two different scoring schemes, much as an expert might do to explore the sensitivity of the calculation to the parameters used. If you supplied a docked ligand, such as a crystallographic ligand, or started from a PDB code, then the ligand is re-docked, and its rank is compared with about 100 property-matched decoys. If you supplied actives or inactives (optional), then these are docked, for comparison with the ligand and its decoys. In every case, two additional databases are always docked. These are extrema and random. Extrema are molecules with calculated LogP < -3 and molecular weight less than 250. They are used to highlight possible weaknesses in docking calculation that would make it highly sensitive to charge. Random are about 200 molecules, quite diverse, that can be used to expose other weaknesses in the docking calculation prior to full database docking.
Calibration can be in one of three states: Queued, Running, and Done.
Your calibration calculation is queued and will start when system resources permit. If the cluster is heavily loaded, you can potentially be stuck in this stage for a long time. We will try to provide timely information to help you manage your expectations.
Your calibration jobs are running. Typically, this should take a few minutes - in unusual cases an hour. If it takes much more than an hour something has probably gone wrong.
Your calibration calculation has ended. At this point, the preliminary docking results should be evaluated with the goal of deciding whether prospective database screening is warranted.
If you supplied a ligand, you will be presented with a Preliminary Docking Results table. Good results are indicated in green, borderline results in yellow, and poor results in red. If the calculation did not complete an uncolored error message will indicate this.
We recommend always looking closely at the ligand and its decoys, which you can do blah blah blah. If you are satisfied with the ligand poses, and if the color is green, then we recommend prospective database screening.
This is when real database docking is done.
Your docking job is queued.
Your docking jobs are running.
Your docking jobs have ended. You may now look at the results.