Any published work which has made use of 3DNA should cite at least one of the following papers:

The current list of journal articles citing 3DNA can be found in Google scholar.

In this section, I will occasionally highlight papers that employ 3DNA in novel or significant ways. If you have a story to share, please let me know.


DSSR-enabled cartoon block images for G-quadruplex

Via Google Scholar, I noticed the recent publication in Nucleic Acids Research by Meier et al. titled Structure and hydrodynamics of a DNA G-quadruplex with a cytosine bulge. Reading through the article, I am pleased to see the section “Nucleic acid geometry and visualization” under MATERIALS AND METHODS:

We used the program DSSR (53) of the 3DNA suite (54) to analyse the nucleic acid backbone and the base pair geometry from the 3D structures. We reported the ‘simple’ base-pair parameters for buckle, propeller twist and stagger which are more intuitive for non-canonical base-pairs than the classic base-pair parameters as explained in the program manual and the 3DNA website (, We wrote an R (55) script that automatically creates a backbone angle plot from the output of the DSSR program. The script can be downloaded from the 3DNA forum at The nucleic acid was visualized in PyMOL and the dssr block plugin (The PyMOL Molecular Graphics Sys- tem, Version 2.0, Schro ̈dinger, LLC, …

This is the first time (I’m aware of) that the ‘simple’ base-pair parameters introduced in 3DNA v2.3 is cited in a peer-reviewed journal article. I’m also glad to know that the blog posts on the 3DNA homepage are read, and even referenced in a publication — which surely will prompt me to write more. This is also the first time that the dssr_block PyMOL plugin is cited. I must say that Figures 1, 5, and 6 from the paper look gorgeous. Among other things, the G-tetrads and the surrounding base identity are immediately obvious using the simple color code: A, red; T, blue, C, yellow, and G, green. See Fig. 1 below.

DSSR-enabled block images [Fig. 1 of Meier18-nar-gky307]

In the section on “DATA AVAILABILITY”, the authors further noted:

Our R (55) script that automatically creates backbone angle plots from the output of the DSSR program can be downloaded from the 3DNA forum at

I communicated with Markus Meier (the lead author) on the 3DNA Forum, on the thread DSSR: Analyzing NMR structures – overwritten output files. Checking the thread right now, I found that the R script (backbone_torsions_plot-1.0.tar.bz2) has been downloaded 263 times. I appreciate Markus’s effort in contributing the R script with a working example to the DSSR user community. It has always been my hope that more DSSR users would share their scripts and examples via the 3DNA Forum.

As a side note, I met Markus in Los Angles at the 60th Annual Meeting. It was a nice experience chatting with him, and had a lunch together. We’ve kept in touch following the meeting.



SNAP is cited in the DNAproDB paper

I recently came across an article titled DNAproDB: an interactive tool for structural analysis of DNA-protein complexes by Sagendorf et al. in Nucleic Acids Research (NAR). Notably, the DNAproDB tool allows users to search the underlying database by combining features of the DNA, protein, or DNA-protein interactions at the interface. Compared to the well-established NUCPLOT tool which generates only ‘static’ schematic diagrams of protein-nucleic acid interactions, DNAproDB is interactive and more user friendly, with many new features.

It was a pleasant surprise to notice that SNAP was cited in the DNAproDB NAR paper, as follows:

Nucleotide-residue interaction geometry (stacking, pseudo-pairing or other) is determined using SNAP, a new component of the 3DNA program suite (35). SNAP also serves as a fall-back for calculating hydrogen bonds if HBPLUS cannot process the file.

I am glad that SNAP has also been used for identifying H-bonds where HBPLUS fails. The H-bonding detection algorithm, initially implemented in 3DNA (v2.3 and before) and refined in DSSR/SNAP, was originally intended to make the 3DNA software fully independent of third-party tools. I did not expect this feature could one day compete with dedicated H-bond finding tools, such as HBPLUS.

By the way, 3DNA is also cited in the DNAproDB NAR paper, as below:

DNA base pairing, shape parameters and conformation are derived from the 3DNA program suite (29) with a 10.0 Å cut-off for helix breaking.



DSSR in the structural analysis of an E.coli initiator tRNAfMet A1-U72 variant

While browsing the latest issue (May 2017) of the RNA journal, I came across the paper titled The structure of an E. coli tRNAfMet A1–U72 variant shows an unusual conformation of the A1–U72 base pair by Monestier et al.. Reading through the text, I am pleasantly surprised by the two references to DSSR as shown below:

An analysis using DSSR (Lu et al. 2015) identifies all the secondary structure elements characteristic of the classical cloverleaf secondary structure as well as usual tertiary interactions that stabilize the L-shaped tertiary fold of the molecule.

As a consequence, the opening parameter (Lu et al. 2015) of the A1–U72 base pair becomes unusually high (153.42°). The NH2 group of A1 points toward the minor groove of the acceptor helix. An interaction between the N1 of A1 and the O2 of U72 (d = 3.0 Å) is observed which requires protonation of the N1 atom of A1.

The PDB id for the deposited structure is 5l4o. Running DSSR on this structure is straightforward: x3dna-dssr -i=5l4o.pdb --more. As with the classic yeast phenylalanine tRNA (PDB id: 1ehz), DSSR identifies two helices, three hairpin loops, one [2,1,5,0] four-way junction loop, among other features.

With regard to the unusual A1-U72 pair highlighted in the title of the paper, DSSR provides the following information. Note the * in the unconventional N1*O2 H-bond.

   1 A.A1           A.U72          A+U --          n/a       tWW  tW+W
       [-14.4(...) ~C3'-endo lambda=32.9] [-172.4(anti) ~C3'-endo lambda=65.0]
       d(C1'-C1')=10.80 d(N1-N9)=9.19 d(C6-C8)=10.68 tor(C1'-N1-N9-C1')=173.6
       H-bonds[1]: "N1*O2(carbonyl)[2.99]"
       interBase-angle=6  Simple-bpParams: Shear=3.53 Stretch=1.71 Buckle=2.0 Propeller=-6.0
       bp-pars: [-0.32   3.91    0.01    6.32    -0.26   153.56]

This citation is yet another example of DSSR’s adoption by experimental biologists. I can only expect to see more such type of DSSR usages in the coming years.



DSSR is used in the URS DataBase

Recently, I came across the article URS DataBase: universe of RNA structures and their motifs by Baulin et al. in Database, an online journal of biological databases and curation. I am glad to see that DSSR is used in the URSDB, as quoted below.

In the “Input data” subsection of “Materials and methods”:

RNA-containing structures were extracted from the PDB in mmCIF format; each file was divided into models. The base pairs (both canonical and non-canonical) and dinucleotide steps were annotated using the DSSR program from 3DNA toolkit (26). We also exploited detailed information provided by DSSR on given elements such as geometric parameters, types according to different classifications and various details on base conformations.

Moreover, under “Future development”, the authors said:

We plan to perform a comparative analysis of programs that annotate base pairs in RNA-containing PDB files. We will consider the four most popular programs, FR3D (35), MC-Annotate (36), RNAView (37) and DSSR (26). According to the analysis the annotation of the base pairs will be refined. In addition, we plan to include in the database annotations of base-phosphate, base-ribose and base stacking contacts and to implement search of such data.

It is gratifying to see DSSR listed as one of “the four most popular programs” for annotating RNA base pairs. It’d also be interesting to see how DSSR compares with FR3D, MC-Annotate, and RNAView from the user’s perspective.



DSSR in the validation of NMR RNA structures

With great interest, I read the article titled Improving NMR Structures of RNA by Bermejo et al. As is well-known, solution NMR structures of RNA normally exhibit more steric clashes and conformational ambiguities than their crystal X-ray counterparts. The paper introduces an improved force field, RNA-ff1, for structure elucidation with Xplor-NIH. By adopting realistic atom radii and a new statistical torsional potential, the RNA-ff1 force field significantly enhances covalent geometry and MolProbity validation scores (in steric contacts and backbone conformation) in the seven tested NMR datasets.

I am glad to see that DSSR is mentioned in the Section titled Analysis of Known Structural Motifs:

… The program DSSR (Lu et al., 2015) (part of the 3DNA software suite [Lu and Olson, 2003, 2008]) was used to evaluate the stacking configuration of successive base pairs (i.e., ‘‘steps’’) within the helical stems of the systems in the present calculations. The most interesting trends are observed for the base-pair step parameters slide (Figure 4K) and rise (Figure 4L), which respectively measure an in-plane dislocation and the vertical displacement of a step relative to a local mid-step reference frame (Lu and Olson, 2003; for analysis of all step parameters, see Figure S1). Relative to A-form parameters in high-resolution X-ray structures (Olson et al., 2001) (Figures 4K and 4L, dashed lines), the average slide of all but one of the original NMR models (PDB: 1O15) is small in absolute value (Figure 4K). … Moreover, four out of the seven original PDB models display an average rise considerably larger than the expected 3.32 Å (the van der Waals separation distance between bases, not to be confused with the helical rise, measured relative to the helical axis, expected to be 2.83 Å for A-form [Olson et al., 2001]).

As an example, the single stem of PDB: 2KOC’s representative structure, assumed to be an A-form helix (Nozinovic et al., 2010), displays a particularly large separation between base pairs C3–G12 and A4–U11 (rise: 4.33 Å) that is visually evident when compared with that of the RNA-ff1 representative model (rise: 3.33 Å ) (Figure 6A). Indeed, this base-pair step defies conformational classification by DSSR in the PDB: 2KOC structure, while it is assigned as A-form (along with the rest of the stem) in the RNA-ff1 structure.

Through the text, the term “stem” or “helical stem” is used consistently, in line with the nomenclatures adopted by DSSR. It is worth noting that DSSR also derives a complete set of backbone conformational parameters, including the assignment of sugar-phosphate backbone suites. The backbone parameters constitutes only a small portion of what DSSR has to offer, and they are written to the auxiliary file dssr-torsions.txt by default.



First mention of SNAP in a peer-reviewed journal article

The other day, I came across an article titled Different duplex/quadruplex junctions determine the properties of anti-thrombin aptamers with mixed folding by Krauss et al. published in Nucleic Acids Research (NAR). This NAR article draw my attention via Google Scholar alert because of its citation to the 2008 3DNA Nature Protocols paper, as shown below (in the Structural analysis section):

3DNA-dssr (41) was used to calculate local and overall geometric parameters of the aptamer. Superpose program from CCP4 package (42) was used to calculate root mean square deviations. Features of the thrombin–RE31 interface were calculated using Cocomaps server (43), whereas contacts between the two molecules, as well as packing interactions between the aptamer and symmetry related thrombin molecules, were found by using 3DNA-snap (41) and Pisa (44) programs. All the results were veri ed by visual inspec- tion of the structure with WinCoot (39).

Moveover, Table 2 lists Stacking interactions as calculated by 3DNA-DSSR (41) among residues belonging to the duplex, the junction and the quadruplex of RE31, with a note on the definition of base-stacking interactions:

Base-stacking is quantified as the area of the overlapped polygon de ned by the two bases of the interacting nucleotides, where the base atoms are projected onto the mean base plane.

To the best of my knowledge, this is the first time SNAP is mentioned in a peer-reviewed journal article. This paper also made good use of DSSR for the analysis of a complicated DNA structure (like RNA), with three non-canonical base pairs at the duplex/quadruplex junction (Figure 3) and extensive stacking interactions (Figure 4).

The duplex/quadruplex junction in RE31 aptamer.
Figure 3. The duplex/quadruplex junction in RE31 aptamer.

Ribbon representation of RE31 highlighting the continuous stacking of bases from the duplex to the quadruplex region.
Figure 4. Ribbon representation of RE31 highlighting the continuous stacking of bases from the duplex to the quadruplex region.

As this paper and those by Paul Paukstelis illustrate, DNA can adopt far more complicated 3D structures enabled by non-canonical base pairing schemes than the simple Watson-Crick paired double helices. 3DNA (including DSSR and SNAP) is well suited for the analysis of such extraordinary structures. On a different perspective, following 3DNA citations has become an effective way for me to keep in pace with relevant literature.



First citation to the DSSR NAR paper in JMB

Recently, I noticed via Google Scholar the first citation to the paper DSSR, an integrated software tool for dissecting the spatial structure of RNA, recently published in Nucleic Acids Research (NAR). The citation is from Srinivas Somarowthu, in a review article titled Progress and current challenges in modeling large RNAs in the Journal of Molecular Biology. The JMB review article is concise, and overall a nice reading.

Specifically, in the section “Model Evaluation and Refinement”, DSSR is listed along with RNAView and MC-Annotate for the characterization of the secondary from 3D atomic coordinates, as below:

After building a model, it is essential to evaluate the quality, find any errors and refine the accordingly. First, it is important to make sure that all the base-pairs and the overall secondary structure is maintained correctly in the model. Tools such as RNAview [82], MC-Annotate [83], and DSSR [84] can calculate the secondary structure from a given 3D structure and thereby allow identification of problematic base-pairs. Recently, Antczak et al [85], developed a web server, RNApdbee, which integrates RNAview, MC-Annotate and DSSR, and extracts not only secondary structures but also kissing-loops and pseudoknots from a target tertiary model. Problematic base pairs can be fixed or rebuilt using interactive tools such as S2S/ASSEMBLE [45].

I am glad to see the first citation to the 2015 DSSR paper per se shortly after its publication in NAR. Looking forward, I can only expect more DSSR citations in diverse fields related to RNA structures.



Citation statistics to 3DNA publications

On October 29, 2015, I performed a survey of citations to the following three 3DNA papers, using the Web of Science. The total number of citations are: NAR03 (787) + NP08 (184) + NAR09 (78) = 1049, spanning a diverse set of 191 journals in biology, chemistry, and material sciences. On the same date, Google Scholar reported 1360 citations for the same three papers.

  1. [NAR03] Lu, Xiang‐Jun, and Wilma K. Olson. “3DNA: a software package for the analysis, rebuilding and visualization of three‐dimensional nucleic acid structures.” Nucleic acids research 31.17 (2003): 5108-5121.
  2. [NP08] Lu, Xiang-Jun, and Wilma K. Olson. “3DNA: a versatile, integrated software system for the analysis, rebuilding and visualization of three-dimensional nucleic-acid structures.” Nature protocols 3.7 (2008): 1213-1227.
  3. [NAR09] Zheng, Guohui, Xiang-Jun Lu, and Wilma K. Olson. “Web 3DNA—a web server for the analysis, reconstruction, and visualization of three-dimensional nucleic-acid structures.” Nucleic acids research 37.suppl 2 (2009): W240-W246.

Among the 1049 citations in 191 journals, 694 citations (66%) are from the following 24 journals (~13%). The remaining 355 citations are from 167 other journals, including Cell (5 times), Science (2), Nature (3) and six additional Nature Publishing Group sub-journals (17).

1 Nucleic Acids Res (167)
2 J Phys Chem B (64)
3 Biochemistry (45)
4 J Am Chem Soc (45)
5 J Mol Biol (41)
6 Phys Chem Chem Phys (28)
7 Biophys J (25)
8 J Biol Chem (23)
9 PLoS One (23)
10 Acta Crystallogr D Biol Crystallogr (22)
11 J Chem Theory Comput (22)
12 Proc Natl Acad Sci U S A (22)
13 Bioinformatics (18)
14 Biopolymers (18)
15 J Biomol Struct Dyn (18)
16 J Chem Phys (18)
17 J Phys Chem A (16)
18 Structure (13)
19 RNA (12)
20 Biochem Biophys Res Commun (11)
21 Chem Res Toxicol (11)
22 J Comput Chem (11)
23 J Mol Model (11)
24 Nat Struct Mol Biol (10)

It is worth noting that while the Web of Science citation report is comprehensive, it is certainly not complete. In particular, citations in the online methods section seem not to be covered. For example, two 3DNA citations (on the DSSR program) in “Materials and Methods” (the Supplementary Materials) of two Science articles by the Ramakrishnan lab are missing from the list. Specifically, the Science papers employed DSSR for the characterization of RNA secondary structural features in crystal structures of the large ribosomal subunit and the whole ribosome of human mitochondria.

For those why are interested in knowing the details, click the link for the full reports of 3DNA citations. In the file, the citations are sorted in two ways: by citation numbers per journal, and by journal names.



The 3DNA mutate_bases program is cited in Nature

It was a nice surprise to notice the following 3DNA citation in a Nature article, titled Selective small-molecule inhibition of an RNA structural element (doi:10.1038/nature15542). Moreover, the work came from Merck Research Laboratories, reporting a novel selective chemical modulator (ribocil) to repress riboswitch-mediated ribB gene expression and inhibit bacterial cell growth.

Homology modelling. A homology model of the E. coli FMN aptamer was constructed using program mutate_bases53 of the 3DNA package using the F. nucleatum impX riboswitch aptamer X-ray structure as the template and the FMN aptamer alignment of E. coli, F. nucleatum, P. aeruginosa and A. baumannii (Extended Data Fig. 5). All nucleotide insertions in the E. coli sequence were removed in the model (Extended Data Fig. 5). There are 34 base changes among the 111 nucleotides modelled. Base pairing when present remains consistent. Energy minimization at A92 was performed to avoid VDW clashes using Macromodel (Schrodinger, LLC).

In retrospect, the mutate_bases program was created in response to repeated requests from 3DNA users, initially mostly for modeling DNA-protein complexes. The program was first coded as a Perl script, and later on rewritten in ANSI C for efficiency. Since v2.1, mutate_bases has become an essential component of 3DNA, on a par with find_pair, analyze, rebuild and fiber etc. As I noted in the post documenting the program

Overall, mutate_bases has been designed to solve the in silico base mutation problem in a practical sense: robust and efficient, getting its job done and then out of the way. The program can have many possible applications: in addition to perform base-pair mutations in DNA-protein complexes, it should also prove handy in RNA modeling and in providing initial structures for QM/MM/MD energy calculations, and in DNA/RNA modeling studies.

The Merck Nature paper is the first time ever that the 3DNA mutate_bases program has been put in the spotlight. Hopefully more such applications/citations will appear in the future as the community begin to appreciate the value of this little gem.



« Older ·

Thank you for printing this article from Please do not forget to visit back for more 3DNA-related information. — Xiang-Jun Lu