From xiangjun@rutchem.rutgers.edu Date: Fri, 6 Jun 2003 01:05:18 -0400 From: Xiangjun LU To: Heinz Sklenar Cc: David Beveridge , Zippi Shakked , Remo Rohs , Thomas Cheatham , Stephen Neidle , Victor Zhurkin , martin.egli , ponzy , Juan.A.Subirana , Karolin.Luger , schlick , loren.williams , e.westhof , Krystyna.Zakrzewska , sponer , m.zacharias , agor , heinemann , olson , haim.rozenberg , Manju Bansal , Richard Lavery , Rama Sarma , Richard E. Dickerson Subject: Re: Albany workshop Dear colleagues: So far, I've found it very interesting to read Dr. Sklenar draft's on "The definition of local nucleic acid parameters" and Drs. Zhurkin and Olson's follow up. I feel I am in a position to provide some technical background about 3DNA that would hopefully clarify some misinterpretations. While trying to resolve the discrepancies among the various nucleic acid analysis programs around 1998-1999, Dr. Olson and I went through all the details to re-implement all the seven commonly used algorithms to make sure that our implementation could exactly reproduce the original parameters, thus to avoid any superficial judgment as to which method is the "best". Indeed, it was found that all the methods would give directly comparable result (e.g. similar patten of parameter variations) if the same reference frame was used, and this is true even for the highly deformed TATA-boxed DNA (except for an offset in Slide, and unusually large Rise for kinked steps, both from Curves). Following the Tsukuba Workshop, we developed 3DNA, which makes use of the "best features" we could combine from the methods mentioned above, including CEHS/SCHNAaP, RNA, CompDNA, NUPARM and NGEOM. Curves has special features that 3DNA does not want to repeat/compete (e.g. global parameters, groove dimension parameters). Nevertheless, we provide an option in a 3DNA utility program (find_pair) to generate input to Curves directly from a PDB data file. It is also worth mentioning that even though the local step parameter (slide/roll) definition in 3DNA mostly follows those in CEHS/SCHNAaP (identical to NGEOM which traced back to Dr. Zhurkin's early work), the numerical values from 3DNA are comparable directly to Curves, rather than to the original CEHS/SCHNAaP parameters. This is simply because the Tsukuba recommended base-centered reference frame matches most closely to that used in Curves (and CompDNA). More specifically, the exact definition of local base-pair parameters used in 3DNA is detailed CEHS/SCHNAaP paper. It uses a symmetric definition, Twist followed by the vector sum of Roll and Tilt, thus avoids the x-then-y or y-then-x ambiguity. This definition can also be "visualized" by following the step-by-step geometric description of each step, and defines the bending angle between two neighbouring base-pairs as sqrt(roll^2+tilt^2). The local helical axis and a set of local helical parameters definition are described briefly in the 3DNA manuscript Dr. Olson sent out several days ago. Overall, it follows exactly the above base-pair parameter definitions, except for a change from mean base-pair normal to local helical axis. The two sets of definition are mathematical rigorous which allows for exact reconstruction from either set (using "rebuild" in 3DNA), and completely reversible thus the simple equations (1) to (4) in our 3DNA manuscript. A commonly asked question is that in 3DNA the "Local base-pair helical parameters" are associated with each dinucleotide step instead of a base-pair. This is because that the helical parameters are *symmetrically* defined that the two base-pairs in the step have *exactly* the same numerical values, and to define the local helical axis (and helical rise and helical twist), two reference frames are required. The set of local helical parameters can be used to discriminate the A-, B- and TA-DNA conformations as discussed in 3DNA manuscript. In summary, 3DNA defines 3 sets of "symmetric" local parameters in a consistent and mathematically rigorous manner: (1) base-pair parameters: Shear, Stretch, Stagger, Buckle, Propeller, Opening; and the "mean" frame is used as the base-pair frame for the calculation of the following two sets of step parameters; (2) base-pair step parameters: Shift, Slide, Rise, Tilt, Roll, Twist and (3) base-pair helical parameters: X-displacement, Y-displacement, helical rise, inclination, tip and helical twist. As mentioned above, in 3DNA, (2) and (3) are inter-convertible, and both can be used to rebuild a DNA structures. I will provde a step-by-step working examle to make the above procedures clear. Best regards, Xiang-Jun