Hi everyone,

I am noticing that Modeller does not do a good job when it comes to sequence alignment of short peptides or DNA sequences. For example, please see the the comparison below:

(1) Modeller alignment:
Template   GTPPPP - - YTVG
   Peptide   TVLPPPPYRHRN

(1) ClustalW alignment:
Template   -GTPPPPYTVG-
   Peptide   TVLPPPPYRHRN

As you can see, Modeller is unable to align up properly the central PPPPY motif in my peptide with the template!! In contrast, ClustalW is spot on in its ability to recognize a key pattern in the sequence.

Is there a way to improve sequence alignment in Modeller? How can I make changes to the following Modeler file that I am using for my sequence alignment:

# Make a pairwise alignment of two sequences
aln = alignment(env, file='template.ali')
aln.salign(overhang=30, gap_penalties_1d=(-450, -50), alignment_type='tree', output='ALIGNMENT')
aln.write(file='template.pir', alignment_format='PIR')
aln.write(file='template.pap', alignment_format='PAP')

What parameters or settings, if any, could be changed in the above setup to improve sequence alignment in Modeller?

Thank you so much for your advice and any comments that you may have.

AMJAD FAROOQ PhD DIC | Associate Professor
Dept of Biochemistry & Molecular Biology | Miller School of Medicine | University of Miami | Miami | FL 33136
Located @ Gautier Building #217 | Mail @ 1011 NW 15th Street #217, Miami , FL 33136
amjad@farooqlab.net | off 305-243-2429 | lab 305-243-9799 | fax 305-243-3955 | www.farooqlab.net