[Dock-fans] Parallel processing resource allocation question

Phan, Jason jason.phan at Vanderbilt.Edu
Wed Apr 18 17:55:22 PDT 2012


Hi John,

Yes, I like to get the structures of the "not actual compound" rings that were extracted from the all purchasable or other subsets. I don't intend to buy them but if they score well and have the right geometry and chemistry to the next pocket, that would be something to explore further. Just heterocyclic rings without any spinach on them is what I want. Of course the risk is they may not exist in the real chemical world without the other attachments but just to dock around, it might provide some insight into the binding pocket.

Thanks,
Jason

On Apr 18, 2012, at 7:13 PM, John Irwin wrote:

> Hi Jason
> 
> Those are not actual compounds, but rings found in compounds. You
> would want to be docking actual purchasable compounds.  You can
> download the compounds in ZINC containing each ring. Is that perhaps
> what you wanted?  You can get that set by clicking on the
> "purchasable" link on the ring detail page.
> 
> Thus all purchasable molecules containing a benzene i.e. phenyl ring
> would be here  http://zinc.docking.org/rings/1/substances/all  It is
> more than 2M compounds and climbing, so this is definitely NOT the
> most efficient way to download them.
> 
> Or explain in other words what you want. (I'm rather slow I'm afraid.)
> 
> John
> 
> 
> On Wed, Apr 18, 2012 at 4:40 PM, Phan, Jason <jason.phan at vanderbilt.edu> wrote:
>> Hi John,
>> 
>> I meant the heterocyclic ring subset on the Zinc.docking.org site. I can use filters to pick them out but if there's a clean subset, it'd be nice to be able to download.
>> 
>> http://zinc.docking.org/browse/rings/?page=1&type=simple
>> 
>> Jason
>> 
>> On Apr 18, 2012, at 6:31 PM, John Irwin wrote:
>> 
>>> Hi Jason
>>> 
>>> What do you mean by the ring subset?
>>> 
>>> John
>>> 
>>> 
>>> On Wed, Apr 18, 2012 at 9:08 AM, Phan, Jason <jason.phan at vanderbilt.edu> wrote:
>>>> Got you! Thanks.
>>>> 
>>>> Is it possible to download the ring subset from ZINC for docking?
>>>> 
>>>> Jason
>>>> 
>>>> On Apr 18, 2012, at 10:53 AM, John Irwin wrote:
>>>> 
>>>>> Jason
>>>>> 
>>>>> My advice to you and anyone using DOCK 5.X or 6.X is to upgrade to
>>>>> DOCK 6.5. It contains many significant changes and bug fixes. It is
>>>>> not worth spending another minute of your life on DOCK5.4.
>>>>> 
>>>>> John
>>>>> 
>>>>> 
>>>>> On Wed, Apr 18, 2012 at 8:46 AM, Phan, Jason <jason.phan at vanderbilt.edu> wrote:
>>>>>> Dear All,
>>>>>> 
>>>>>> Sorry if this has been addressed already and if so please point me to those threads. I'll move to dock6.5 soon but right now, I'm stuck with dock5.4. Using mpiexec -n 12 dock5mpi -i etc..., it only lets me use 12 processors (e.g. nodes=4:ppn=3 or nodes=6:ppn=2 or any of the combinations that add up to 12 processors total). If I go over 12, it got bumped out with this error: "Error: Missing parameter/value pair" and eventually, "Parameterization Failed. 16 missing parameters". If I stay at 12 or lower, no problem. Also, while this is running, I cannot start another dock run... same error. It looks like these processes are competing for some parameter files. I did copy them to a local directory but did not help. Any suggestions?
>>>>>> 
>>>>>> Thanks,
>>>>>> Jason
>>>>>> _______________________________________________
>>>>>> Dock-fans mailing list
>>>>>> Dock-fans at docking.org
>>>>>> http://mailman.docking.org/mailman/listinfo/dock-fans
>>>>> 
>>>> 
>>>> 
>>>> _______________________________________________
>>>> Dock-fans mailing list
>>>> Dock-fans at docking.org
>>>> http://mailman.docking.org/mailman/listinfo/dock-fans
>>> 
>> 
>> 
> 




More information about the Dock-fans mailing list