Re: ABF - using gyration

From: Subramanian Vaitheeswaran (vaithee_at_umd.edu)
Date: Tue Oct 28 2008 - 15:19:55 CDT

Jerome,

I'll try out these suggestions and let you know the result. About the number of atoms in the rg calculation - I am using all peptide atoms, perhaps I could use only carbons. Using only alpha carbons will probably be not be a good idea since I have large side chains like tryptophan and tyrosine. Thanks a lot for the quick responses.

Vaithee
============================================
S. Vaitheeswaran
Post-doctoral Research Associate
Institute for Physical Science & Technology
Rm 2116, Bldg. 85
University of Maryland
College Park MD 20742
email: vaithee_at_umd.edu
Phone: 301-405-7568
============================================

---- Original message ----
>Date: Tue, 28 Oct 2008 16:11:43 -0400
>From: "Jerome Henin" <jhenin_at_cmm.chem.upenn.edu>
>Subject: Re: namd-l: ABF - using gyration
>To: "Subramanian Vaitheeswaran" <vaithee_at_umd.edu>
>Cc: namdlist <namd-l_at_ks.uiuc.edu>
>
>Vaithee,
>I think you're running into a problem with the way the code does two things:
>- enforce the restraining potentials
>- apply the Jacobian (geometric) contribution to the PMF.
>In the code you have, a Jacobian correction is also added on top of
>the harmonic restraining potential. In the case of a radius of
>gyration with many atoms, this results in a strongly negative force
>that interferes with the harmonic restraint and gives a local minimum
>outside the sampling grid.
>
>Attached is a variant of the scripts that should not suffer from this
>problem. You could try increasing the strength of the boundary
>potential as well. Also, it may not be useful to include so many
>atoms: the radius of gyration of a subset of atoms is likely to give
>you the same information. Please let me know if the problem persists
>with these changes.
>
>Last, the development (CVS) version of NAMD contains an altogether new
>codebase for ABF, with better performance and many additional
>features. Consider trying it out if you have the time.
>
>Best,
>Jerome
>
>On Tue, Oct 28, 2008 at 3:19 PM, Subramanian Vaitheeswaran
><vaithee_at_umd.edu> wrote:
>> Jerome,
>>
>> I am attaching two figures. The first is a comparison of the PMFs from the unbiased and ABF runs. The second figure shows the rg values for these same data sets. I don't understand why the sampling is so far off the set range.
>>
>> thanks,
>> Vaithee
>> ============================================
>> S. Vaitheeswaran
>> Post-doctoral Research Associate
>> Institute for Physical Science & Technology
>> Rm 2116, Bldg. 85
>> University of Maryland
>> College Park MD 20742
>> email: vaithee_at_umd.edu
>> Phone: 301-405-7568
>> ============================================
>>
>>
>> ---- Original message ----
>>>Date: Tue, 28 Oct 2008 12:52:49 -0400
>>>From: "Jerome Henin" <jhenin_at_cmm.chem.upenn.edu>
>>>Subject: Re: namd-l: ABF - using gyration
>>>To: "Subramanian Vaitheeswaran" <vaithee_at_umd.edu>
>>>Cc: namdlist <namd-l_at_ks.uiuc.edu>
>>>
>>>Hi Vaithee,
>>>Can you compute a local PMF around the minimum from your unbiased
>>>simulation, and compare it with what the ABF runs measure?
>>>Best,
>>>Jerome
>>>
>>>On Tue, Oct 28, 2008 at 11:18 AM, Subramanian Vaitheeswaran
>>><vaithee_at_umd.edu> wrote:
>>>> Dear Chris and Jerome,
>>>>
>>>> I am trying to use ABF to calculate the PMF of a peptide with the radius of gyration as the order parameter. But my system is behaving strangely.
>>>>
>>>> I first ran the system (a decapeptide in a box of water) in NAMD *without* any biasing and measured the radius of gyration, rg to be in the range [6.0,7.5]. I then used ABF with "coordinate gyration" with the same range, expecting to see sampling in this region. But xi, the order parameter, rapidly drops to ~5.6.
>>>>
>>>> Next, I reduced range to [5.4,5.8] - then xi drops to ~5.4. So, the biasing seems to be driving the system opposite to the desired direction. This is the ABF part of my conf file:
>>>> **************************
>>>> if {1} {
>>>> source ~/NAMD_2.6_Linux-i686/lib/abf/abf.tcl
>>>> abf coordinate gyration
>>>> abf abf1 {1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169}
>>>> abf dxi 0.05
>>>> abf xiMin 5.4
>>>> abf xiMax 5.8
>>>> abf writeXiFreq 5000
>>>> abf outFile rg.pmf
>>>> abf historyFile rg.hist
>>>> abf fullSamples 200
>>>> abf inFiles {./1_5ns/rg.pmf}
>>>> abf dSmooth 0.0
>>>> }
>>>> **************************
>>>> Both ABF runs were ~2ns long at a 2fs time step. What could be going wrong?
>>>>
>>>> thanks,
>>>> Vaithee
>>>> ============================================
>>>> S. Vaitheeswaran
>>>> Post-doctoral Research Associate
>>>> Institute for Physical Science & Technology
>>>> Rm 2116, Bldg. 85
>>>> University of Maryland
>>>> College Park MD 20742
>>>> email: vaithee_at_umd.edu
>>>> Phone: 301-405-7568
>>>> ============================================
>>>>
>>>>
>>>
>>
>________________
>abf_script.tcl (30k bytes)
>________________
>gyration.tcl (4k bytes)

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:21:25 CST