Hi Julianna,
Thanks for your follow-up. In addition to serving your purpose of extending DNA from both ends, the PDB entry
1da0 is a good example to help illustrate one subtle point in 3DNA.
By default, 3DNA handles one structure at a time. When given an ensemble (as in NMR) with multiple structures in MODEL/ENDMDL delineated PDB format, 3DNA only analyzes the first model. Since 1da0 is symmetrical, with only one chain (and one daunomycin) in the asymmetric unit, you should download the coordinates in biological unit. The file name would be 1da0.pdb1 (the first and only biological assembly), where the two symmetric-related chains are separated by MODEL/ENDMDL pairs:
HETATM 158 O5' DM1 A 7 -12.846 8.039 6.403 1.00 23.07 O
HETATM 159 C6' DM1 A 7 -14.267 6.866 7.856 1.00 18.53 C
ENDMDL
MODEL 2
ATOM 1 O5' DC A 1 -20.437 18.629 3.262 1.00 14.14 O
ATOM 2 C5' DC A 1 -21.455 17.960 2.453 1.00 8.69 C
To proceed with 3DNA, make the ENDMDL record defective by adding a '#' (or a space) in front of it:
#ENDMDLLet's call the modified file 1da0-biou.pdb (attached), which is the starting point of following steps.
First, run
find_pair to identify base pairs (bps), and generate the reference frame file (ref_frames.dat):
1da0-biou.pdb
1da0-biou.out
2 # duplex
6 # number of base-pairs
1 1 # explicit bp numbering/hetero atoms
1 13 0 # 1 | ...1>A:...1_:[.DC]C-----G[.DG]:...6_:A<...2 0.47 0.02 5.52 8.64 -4.21
2 12 0 # 2 | ...1>A:...2_:[.DG]G-----C[.DC]:...5_:A<...2 0.45 0.20 11.07 8.54 -1.60
3 11 0 # 3 | ...1>A:...3_:[.DA]A-----T[.DT]:...4_:A<...2 0.75 0.01 8.35 8.62 -3.81
4 10 0 # 4 | ...1>A:...4_:[.DT]T-----A[.DA]:...3_:A<...2 0.75 0.01 8.35 8.62 -3.81
5 9 0 # 5 | ...1>A:...5_:[.DC]C-----G[.DG]:...2_:A<...2 0.45 0.20 11.07 8.54 -1.60
6 8 0 # 6 | ...1>A:...6_:[.DG]G-----C[.DC]:...1_:A<...2 0.47 0.02 5.52 8.64 -4.21
##### Base-pair criteria used: 4.00 0.00 15.00 2.50 65.00 4.50 7.80 [ O N]
##### 0 non-Watson-Crick base-pairs, and 1 helix (0 isolated bps)
##### Helix #1 (6): 1 - 6To extend the structure at the 3' end, run:
frame_mol -6 ref_frames.dat 1da0-biou.pdb 1da0-biou-ref6.pdbto get file 1da0-biou-ref6.pdb where the structure is in the reference frame of bp #6 (i.e., the last bp).
Then, use the
fiber command to generate a B-DNA duplex of your preferred sequence:
fiber -seq=AATTTCCGGG frag3.pdbHere the sequence is AATTTCCGGG, and the resultant duplex is named frag3.pdb.
Now set frag3.pdb in the reference frame of the first bp, and name the resultant file frag3-ref1.pdb:
find_pair frag3.pdb stdout
frame_mol -1 ref_frames.dat frag3.pdb frag3-ref1.pdbNow you can combine structures 1da0-biou-ref6.pdb and frag3-ref1.pdb, since they share a common reference frame, into a file named 1da0-biou-extend3.pdb (attached):
comb_str 1da0-biou-ref6.pdb frag3-ref1.pdb 1da0-biou-extend3.pdbDisplay 1da0-biou-extend3.pdb using Jmol/PyMOL, you will see that 1da0 has been extended in its 3'-end (see figure below). Note the overlap of atoms at the joint bp. You need to remove the redundant part to get a sensible DNA structure.
The method outlined here may look complicated, but it is essential to get a clear understand of the underlying principles. Once passing through the hurdle, it won't be that difficult to automate the whole process to achieve what you want efficiently and reproducibly.
You can follow a similar procedure to extend 1da0 in the 5'-end. It would help if you work out the details and share back with the community.
HTH,
Xiang-Jun
Show Posts
. As noted in my previous reply, this bug did not show up on Mac OS X either. So in the future bug report, please remember to give sufficient details. 
when clicking directly on the topic subject. [Note added on 2014-01-10: the broken URL bug has been fixed.]