[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [modeller_usage] problems modeling a multimer



Douglas Kojetin wrote:
I would like create a hexameric model using a target sequence and a hexameric PDB template. I am trying to create an alignment using input files that have the \ character to separate the chains. Example:

 >P1;target
sequence:target:::::::0.00: 0.00
AAAAA\
BBBBB\
CCCCC\
DDDDD\
EEEEE\
FFFFF*

Do you actually mean the / character here? '/' designates a chain break, but '\' has no special meaning in Modeller alignments.

Using modeller9v1 (linux RPM), I started the above calculation (python align2d.py), but it has not finished after 20 hours

I suggest you try setting the max_gap_length argument to align2d to something smaller than the default - say 20. By default it will backtrack through your alignment matrix to try and find the very best alignment, and for long sequences this can be extremely time consuming, so it is best to limit the gap length.

according to 'top'). Although I used log.verbose(), there is very little output in the associated log (see below).

Yes - you won't see any output from align2d until it has finished.

You may also want to try aligning each chain individually. There is no clear way to align chain breaks, so most of Modeller's alignment routines will either remove them entirely or simply ignore them. In many cases you want the chain breaks in your templates to be "aligned" with the chain breaks in the model, and one way to do that is to make separate alignments for each chain, and then recombine them later.

	Ben Webb, Modeller Caretaker
--
             http://www.salilab.org/modeller/
Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage