Sequence Docking and Side Chain Building (#4)

Given a reasonable main chain trace (the output file autobuild_main.brk from warpNtrace.sh) and the sequence of the protein, the automated script side_dock.sh will attempt to match these main chain fragments to the sequence and build the side chains. It will also create an O macro for building the rest of the structure (but unfortunately manually). The sequence information should be given in a so-called PIR file which can look like the following (all these lines are in a valid format, but your file does not necessarily have to be such a mess) :

>psp

VVRDVNWGALRIAVSTEDLTDPAYHCARVG
LIPQAVQLHT ERLKVQQVQG KWKVTDMVGD ICGDFKVPQA HITEGFSNTD FVMYVASVPS
E E G V L A W A T T CQTFSDGHPA

The only restriction is that the first line must start with the > symbol and the second line should be blank. The rest is worked out by the script.

Tricks and Tips

• If you have a dimer or a higher multimer things get more complicated. It is not easy for side_dock.sh to understand which chain belongs to which monomer by itself. What you should do, is go briefly to the graphics, decide which chain belongs to which monomer and then run side_dock.sh with the chain names of the first monomer as command line arguments: (i.e. side_dock.sh test A B C) and then for the second monomer, (i.e. side_dock.sh test D E F G), etc. That will output several files in the files directory with rather obvious names.

• Take a look at the average confidence level. If its too low, below 10-15, it means that something might well be wrong. It basically means that the program can not decide for sure where to place some fragments into the sequence (there are good enough alternatives) and the side chain docking is not to be trusted too much! Examine the log file (logs/side_dock.log) and see if all the fragments or only a few cannot be placed accurately. Based on that you could decide what to trust and what not.