Hi,
Yes refmac, use weighting term 0.05
Thanks
On Thu, 1 Mar 2007, Tim Gruene wrote:
> How did you do the refinement? Did you use refmac? Did you let the matrix value at 0.5? That's too high. lower it to, say, 0.1 to start with.
>
>
>
> -------- Original-Nachricht --------
> Datum: Thu, 1 Mar 2007 03:51:50 -0500
> Von: Yanming Zhang <[log in to unmask]>
> An: [log in to unmask]
> CC:
> Betreff: [ccp4bb] Very weird
>
>> Dear all,
>>
>> I am sorry to trouble you again because I am facing a very weird
>> situation:
>>
>> Three copies from Phaser are the right solutions based upon:
>> 1, Rfree 42% R 39%
>> 2, No packing clash
>> 3, The packing within the 3 makes good sense
>> 4, Density evenly distributed among the 3 copies even without NCS
>>
>> Then fix these 3, asking Phaser to find one more (here I adopt the
>> strategy one copy at a time). Phaser did find one more, So now I have 4
>> copies in total.
>>
>> Using the phase calculated from the previous 3 copies(because Rfree 42%),
>> the 4th copy can exactly match one junk of the density, even the side
>> chain matchs, demonstrating the 4th copy is correct.
>>
>> Checking the crystallographic packing, The density which the 4th occupies,
>> is an extra density which did not account for by the previous 3 copies or
>> their symmetry mates.
>>
>> No Clash exists
>>
>> Then I use the 4 copies to do the refinement, forseeing that there will be
>> a huge Rfree drop from 42% because I account for the extra density using
>> the 4th copy. BUT TO MY SURPRISE, THE RFREE INCREASE TO 65%.
>>
>> Can you teach me what is going wrong? Thank you
>> Yanming
>
> --
> "Feel free" - 10 GB Mailbox, 100 FreeSMS/Monat ...
> Jetzt GMX TopMail testen: www.gmx.net/de/go/mailfooter/topmail-out
>
|