[modeller_usage] Re: Generating chimeric protein from templates

[Modeller Caretaker <>]

On 2/3/23 9:54 PM, Ta-Chou Huang via modeller_usage wrote:
> However, using the build_profile.py, it shows the first error:
> _modeller.FileFormatError: parse_pir__E> Invalid PIR file header line: 
> structureN:2xkm
> There should be 10 fields separated by colons, :
> This line actually contains 2 fields.

This is correct - you need to provide information on the structure so 
that Modeller knows which PDB to read, and which range of residues in 
that PDB file match the sequence. The alignment file format is 
documented at https://salilab.org/modeller/10.4/manual/node501.html

> Now here comes my question, from the FAQ, I thought it was enough to use 
> two fields for the template?

I had thought it was clear that the alignment file in the FAQ was not a 
"real" alignment (obviously the sequence isn't really "aaaa" either!) 
but I will try to clarify the text.

> with the same build_profile.py, the third error comes:
> _modeller.StatisticsError: regress_657E> Not enough bins in histogram - 
> cannot calculate statistics, nbins:        1

The purpose of profile.build() is to build a profile by scanning your 
sequence against a large database of sequences - usually something like 
PDB95 or UniProt90. This error suggests that you have given it a much 
smaller database, perhaps even a single sequence.

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