AMBER Archive (2009)

Subject: Re: [AMBER] the problem of nucgen.dat

From: case (
Date: Sun Nov 08 2009 - 10:19:03 CST

On Sat, Nov 07, 2009, wrote:
> linedifinition atom no. atom name residue name residue no.
> >>x y z
> That is the format in the manual of Amber10.

Can you point us to exactly which page or section you are getting this
information from? We may need to fix the manual, but I don't see what is
confusing you. [If you are instead looking at section 2.1.1 of tutorial B1,
the file being described there (as above) is nuc.pdb, *not* nucgen.dat.]

[Note to amber developers: tutorial B1 is becoming quite long in the tooth,
and needs an updating, e.g. to use the pdb version 3 atom and residue names.
Volunteers are welcome.]

> We were trying yo run a tutorial.According to its steps,we want to
> use nucgen.dat to build a PDB file for a nucleic acid.This is method
> is similar to homolog modeling.But the final PDB we got is empty.

This is why I asked for more details about exactly what you did. I'm assuming
you are running tutorial B1, but that works fine for me. So, you should
give information about what compiler and OS you are using; what value of
$AMBERHOME you have, and exactly the commands that you used. Did you look
at the nucgen.out file? That might often contain information when something
goes wrong.

In any event, since the tutorial provides you with the output nuc.pdb file,
you can continue with the Amber part of the tutorial, even if nucgen is
failing for you. As I pointed out earlier, nucgen is way outdated, and there
are other good ways to get initial structures for duplex nucleic acids.


AMBER mailing list