Automation of making nucleic acid helices by NAB.
8/03/13: Fixed naming of terminal hydrogens. Thanks to Dr. Rajesh Raju for bringing this to my attention.
1/28/12: Abasic residues are now supported. A "_" indicates an abasic site in the ASCII-art. The structure file will name the abasic as "3DR" for DNA strands and " N" for RNA, according to Protein Data Bank conventions. Abasics are discussed in the instructions and appear in an example below. In addition to adding support for abasic residues, I have corrected some glaring errors in the source code instructions. Finally, I have fixed a very subtle bug that caused a "could not coerce" error when the user desired only a single strand and the ascii art had empty lines following the ascii art sequence.
8/18/11: I have added the option to create only one strand. Just enter the sequence of the strand you want (5' -> 3') and check the Single Stranded checkbutton in the Advanced Options. Beware: if you specify two strands and have checked Single Stranded, then you'll only get back the top strand.
6/11/10: It is now possible to specify only one strand and to have the complementary strand "guessed" by the server. The rule is that if any U is in the strand given, the complement will be constructed as RNA. If a heteroduplex is desired or if you aren't getting the results you expect, just explicitly specify both strands.
1/12/10: The core make-na code has always been OSS, but now the source is available online for direct download. The license has been updated to the GPLv3.
5/14/09: Fixed missing terminal hydrogens for underhanging strands. Thanks to Chongbo Sun at the University of Alberta for bringing this to my attention.
Using the make-na server should be intuitive. ASCII-art is used to depict duplexes and the server tries its best to infer what is meant by the depiction.
I have tried to put some flexibility in duplex creation. The server will create A-, B, and SB-DNA as well as A- and B-RNA. Z-DNA and Z-RNA are not supported yet. If someone wants it, I'll try to add it. Mismatches are supported as well as uneven over-hangs and under-hangs. Abasic residues are also supported, indicated by a "_".
The structure file represents the abasic as the 3-letter code "3DR" in DNA strands and " N" in RNA strands. These are Protein Data Bank conventions.
Following are several valid ASCII-art duplexes:
ATACCGATACGATAGAC TGGCTATGCTATCTGTA CGATAC__TAGAC uGGCuAuGCuAuCuGuA ACTATAGGATACCAT GATAggCTATGGT_
Moreover, the program does its best to ignore comments and other mark-up, so the following will work just fine (the alignment of the two strands will be interpreted correctly):
# This is my favorite DNA. 5' - ATACCGATACGATAGAC -3' 3' - TGGCTATGCTATCTGTA -5'
Comments start with a "#" (preceeded by optional whitespace) and all non-alphabetic information is ignored on sequence lines, except for calculating alignment of the strands. Blank and whitespace only lines are ignored.
It is not necessary to specifiy both strands as a super-shorthand:
# This is my favorite blunt DNA. 5' - ATACCGATACGATAGAC -3'
If the one-strand "super-shorthand" is being used, then a blunt-ended canonical Watson-Crick helix will be produced. Therefore, this latter example is equivalent to:
5' - ATACCGATACGATAGAC -3' 3' - TATGGCTATGCTATCTG -5'
If an abasic is specified using super-shorthand, then an abasic will be put in the partner postion in the complement strand. If any U is in the strand provided, then the complement will be generated as RNA. If a heteroduplex is desired, then it is important to specify both strands explicitly. Remember, the parser is guessing what the input should be, so if the input should be anything beyond the simplest case, then the input should be specified explicitly.
The make-na text parser and modeling code is open source
and available under the
Click here for a source tar file as of 1/28/2012. Instructions for the usage of the source are within.
I've been picked on just a little for these models. To discourage additional criticism, I issue the following disclaimer:
These models may deviate from ideality (whatever ideality really is). I built this server to facilitate the making of models for downstream use in crystallography, NMR, and perhaps molecular dynamics. The pure A-, B-, and SB- forms with canonical WC pairing are unmodified from the NAB fiber-diffraction forms as created by the fd_helix function. See the NAB manual for details. I may apply some limited regularization by molecular dynamics to non-standard models or unusal helix types like Z-DNA. In all cases, take these models with a grain of salt. You will want to apply your own molecular dynamics or constraints from experimental data before you make conclusions from these models. I am not responsible for any misfortune that may befall you if you use these models. Use these models at your own risk.
Having solved several structures containing DNA, I can say with some authority that much variation in nucleic acid structure exists. Considering this variability, please realize that these models are actually very conservative.
A- and B-RNA and DNA parameters and SB-DNA parameters (the "SB" option) come from fiber diffraction data as described in the NAB manual. Here are the references for these types, ripped straight from the manual:
If you have any suggestions or comments, please use my contact page.