Hi Teresa,
Decreasing the temperature that much is probably a bad thing - it will decrease your calculations' ability to escape local energy minima. If you suffer the "calculations blowing up"/atoms with no coordinates problem, it's better to decrease the timestep in the dynamics, or change the random number seed instead. Alternatively look at the class of restraints that are causing the problem and consider the changing the relative weighting of the different types of restraints.
Increasing the number of steps in the dynamics (double, quadruple) is usually a good idea in ARIA calculations of anything but the smallest proteins.
Brian
________________________________________
From: CcpNmr software mailing list [[log in to unmask]] on behalf of Teresa Almeida [[log in to unmask]]
Sent: 09 October 2017 23:06
To: [log in to unmask]
Subject: Re: Using a template structure in Aria2
Hi Nick,
Not too sure if the same applies to you, but when doing structure
calculations I used one of the structures (.pdb) previously obtained
from ARIA calculations hence the formatting was right and I can't
discard the possibility that there may be something wrong with the
structure you uploaded.
However and despite that the calculations failed to run initially and
decreasing the temperature from 10000 to 2000, and increasing the number
of cooling steps (see figure attached for detailed SA parameters) seemed
to work.
Hope this helps.
---
Kind regards
Teresa Almeida
Postdoctoral Research Associate
University of Cambridge
Department of Biochemistry
Cambridge
CB2 1GA
United Kingdom
+44 122361348
On 2017-10-09 13:18, Darvill, Nick wrote:
> Dear all,
>
> Has anyone had any experience using a homology model .pdb as a
> template structure for a structural calculation
>
> My project setup in Aria2 is as follows:
>
> <!DOCTYPE project SYSTEM "project1.0.dtd">
> <project name="tgph" version="1.0" author="" date="Fri Oct 6 18:14:33
> 2017" description="" comment="" references=""
> working_directory="/nmr/aria/nd914/TgAPH800Hz"
> temp_root="/nmr/aria/nd914/TgAPH800Hz" run="12"
> file_root="tgaph_homology" cache="yes" cleanup="yes">
> <data>
> <ccpn_model filename=""/>
> <molecule
> file="/nmr/aria/nd914/TgAPH800Hz/tgph_minusresi1to10_seqout.xml"
> format="xml" ccpn_id="">
> <linkage_definition name="automatic" filename=""/>
> <parameter_definition name="automatic" filename=""/>
> <topology_definition name="automatic" filename=""/>
> </molecule>
> <spectrum enabled="yes" use_assignments="yes"
> trust_assigned_peaks="no" structural_rules="no"
> filter_diagonal_peaks="yes" filter_unassigned_peaks="no">
> <shifts
> file="/nmr/aria/nd914/TgAPH800Hz/tgphcnoesy_ppm_minusresi1to10.xml"
> format="xml" ccpn_id="" default_shift_error="0.0"/>
> <peaks file="/nmr/aria/nd914/TgAPH800Hz/cnoesy_edit_6_cd_xpk.xml"
> format="xml" ccpn_id="" peak_size="intensity"
> freq_window_proton1="0.05" freq_window_hetero1="0.4"
> freq_window_proton2="0.05" freq_window_hetero2="0.4">
> <lower_bound_correction value="0.0" enabled="no"/>
> <upper_bound_correction value="6.0" enabled="yes"/>
> </peaks>
> <experiment_data molecule_correlation_time="10.0"
> spectrum_mixing_time="100.0" spectrometer_frequency="800.0"
> ambiguity_type="all"/>
> </spectrum>
> <spectrum enabled="yes" use_assignments="yes"
> trust_assigned_peaks="no" structural_rules="no"
> filter_diagonal_peaks="yes" filter_unassigned_peaks="no">
> <shifts
> file="/nmr/aria/nd914/TgAPH800Hz/tgphnnoesy_ppm_minusresi1to10.xml"
> format="xml" ccpn_id="" default_shift_error="0.0"/>
> <peaks file="/nmr/aria/nd914/TgAPH800Hz/nnoesy_edit_5_cd_xpk.xml"
> format="xml" ccpn_id="" peak_size="intensity"
> freq_window_proton1="0.05" freq_window_hetero1="0.5"
> freq_window_proton2="0.05" freq_window_hetero2="0.5">
> <lower_bound_correction value="0.0" enabled="no"/>
> <upper_bound_correction value="6.0" enabled="yes"/>
> </peaks>
> <experiment_data molecule_correlation_time="10.0"
> spectrum_mixing_time="100.0" spectrometer_frequency="800.0"
> ambiguity_type="all"/>
> </spectrum>
> <unambiguous_distance_restraints file="/nmr/aria/bliu/gp5.7/all.tbl"
> format="tbl" ccpn_id="" enabled="no" add_to_network="no"
> calibrate="no" run_network_anchoring="no" filter_contributions="no"/>
> <hbonds file="" format="tbl" ccpn_id="" enabled="no"
> data_type="standard"/>
> <dihedrals file="/nmr/aria/nd914/TgAPH800Hz/dihedral_minus1to10.tbl"
> format="tbl" ccpn_id="" enabled="yes" data_type="standard"/>
> <symmetry enabled="no" method="standard" n_monomers="1"
> symmetry_type="None" ncs_enabled="no" packing_enabled="no"/>
> <initial_structure file="/nmr/aria/nd914/TgAPH800Hz/TgAPH_model.pdb"
> format="iupac" ccpn_id="" enabled="yes"/>
> </data>
>
> As you can see I have initial structure enabled using an IUPAC format
> but when I look in CNS>begin for the Aria run, the template structure
> is there (TgAPH_model.pdb) but another template structure (just a
> linear structure) has been generated suggesting Aria2 is recognising
> that I want to use the template structure but is not actually using
> it.
>
> I'm relatively new to using Aria2, so any help would be greatly
> appreciated,
>
> Cheers
>
> Nick Darvill
|