As of v2.5.4-2025jun06, DSSR automatically checks for steric clashes or exact duplicates of residues in an input coordinate file. It reports such issues instead of crashing, and will terminate only if an excessive number of overlaps are detected. An simplified example is shown below, which contains two nucleotides (G#1) on chains 0 and 1, respectively
ATOM 1 OP3 G 0 1 -4.270 51.892 37.186 1.00 27.93 O
ATOM 2 P G 0 1 -3.834 50.887 37.436 1.00 28.61 P
ATOM 3 OP1 G 0 1 -4.601 49.700 37.549 1.00 27.02 O
ATOM 4 OP2 G 0 1 -4.061 52.011 36.684 1.00 25.80 O
ATOM 5 O5' G 0 1 -2.906 51.105 38.691 1.00 28.01 O
ATOM 6 C5' G 0 1 -1.941 52.126 38.781 1.00 26.76 C
ATOM 7 C4' G 0 1 -1.037 51.914 39.967 1.00 26.12 C
ATOM 8 O4' G 0 1 -1.822 51.894 41.184 1.00 24.21 O
ATOM 9 C3' G 0 1 -0.285 50.591 39.988 1.00 25.12 C
ATOM 10 O3' G 0 1 0.884 50.614 39.172 1.00 26.09 O
ATOM 11 C2' G 0 1 0.008 50.411 41.462 1.00 26.05 C
ATOM 12 O2' G 0 1 1.102 51.209 41.880 1.00 27.46 O
ATOM 13 C1' G 0 1 -1.271 50.952 42.083 1.00 28.40 C
ATOM 14 N9 G 0 1 -2.272 49.904 42.329 1.00 27.27 N
ATOM 15 C8 G 0 1 -3.470 49.733 41.686 1.00 26.55 C
ATOM 16 N7 G 0 1 -4.137 48.712 42.125 1.00 25.36 N
ATOM 17 C5 G 0 1 -3.332 48.176 43.118 1.00 25.64 C
ATOM 18 C6 G 0 1 -3.529 47.056 43.955 1.00 24.98 C
ATOM 19 O6 G 0 1 -4.492 46.284 43.991 1.00 24.56 O
ATOM 20 N1 G 0 1 -2.460 46.862 44.821 1.00 24.78 N
ATOM 21 C2 G 0 1 -1.346 47.639 44.878 1.00 24.96 C
ATOM 22 N2 G 0 1 -0.417 47.298 45.782 1.00 23.72 N
ATOM 23 N3 G 0 1 -1.145 48.689 44.109 1.00 25.74 N
ATOM 24 C4 G 0 1 -2.171 48.901 43.257 1.00 26.32 C
ATOM 1 OP3 G 1 1 -6.437 51.060 40.254 1.00 27.81 O
ATOM 2 P G 1 1 -5.327 50.209 39.884 1.00 28.55 P
ATOM 3 OP1 G 1 1 -5.668 48.792 39.652 1.00 26.90 O
ATOM 4 OP2 G 1 1 -4.838 51.036 38.808 1.00 25.57 O
ATOM 5 O5' G 1 1 -4.301 50.297 41.090 1.00 27.94 O
ATOM 6 C5' G 1 1 -3.427 51.393 41.257 1.00 26.67 C
ATOM 7 C4' G 1 1 -2.528 51.168 42.443 1.00 26.12 C
ATOM 8 O4' G 1 1 -3.335 50.964 43.624 1.00 24.16 O
ATOM 9 C3' G 1 1 -1.648 49.928 42.372 1.00 25.13 C
ATOM 10 O3' G 1 1 -0.467 50.136 41.599 1.00 26.15 O
ATOM 11 C2' G 1 1 -1.372 49.649 43.835 1.00 25.96 C
ATOM 12 O2' G 1 1 -0.375 50.515 44.354 1.00 27.37 O
ATOM 13 C1' G 1 1 -2.714 50.006 44.458 1.00 28.21 C
ATOM 14 N9 G 1 1 -3.608 48.845 44.581 1.00 27.06 N
ATOM 15 C8 G 1 1 -4.771 48.614 43.895 1.00 26.37 C
ATOM 16 N7 G 1 1 -5.340 47.496 44.226 1.00 25.18 N
ATOM 17 C5 G 1 1 -4.502 46.957 45.190 1.00 25.44 C
ATOM 18 C6 G 1 1 -4.599 45.755 45.923 1.00 24.77 C
ATOM 19 O6 G 1 1 -5.480 44.892 45.864 1.00 24.39 O
ATOM 20 N1 G 1 1 -3.532 45.594 46.796 1.00 24.63 N
ATOM 21 C2 G 1 1 -2.504 46.469 46.949 1.00 24.81 C
ATOM 22 N2 G 1 1 -1.560 46.145 47.845 1.00 23.58 N
ATOM 23 N3 G 1 1 -2.396 47.594 46.280 1.00 25.56 N
ATOM 24 C4 G 1 1 -3.422 47.779 45.423 1.00 26.12 C Running DSSR on the above coordinates will show the following output:
[i] 0.G1 and 1.G1 in clashes: min_dist=0.57
where min_dist refers to the minimum distance between heavy atoms of the two nucleotides.
The clash-detection feature in DSSR was added in response to the bioRxiv preprint by Kretsch et al. (2025), titled "Assessment of nucleic acid structure prediction in CASP16" (https://doi.org/10.1101/2025.05.06.652459), which noted that in some predicted RNA models submitted to CASP16, multiple models were not properly delineated with MODEL/ENDMDL in PDB format or _atom_site.pdbx_PDB_model_num in mmCIF format. I communicated with the authors, who kindly provided the PDB files to help debug the issue. For more details, see the blog post Improving DSSR through extreme cases from early June 2025 at https://home.x3dna.org/highlights/improving-dssr-through-extreme-cases.
The bioRxiv paper by Kretsch et al. was recently published in Proteins: Structure, Function, and Bioinformatics. The relevant citation to DSSR is in Section 2.8 | Secondary Structure Analysis, as follows:
Secondary structures were extracted from CASP16 models with DSSR (v1.9.9-2020feb06) [47]. Some models, in particular due to large clashes, could not be processed by DSSR (Table S1). The base-pair list was extracted from the table in the output file directly because the dot-bracket structure produced by DSSR, in particular for multimers, contained errors. The canonical base pairs were defined as those labeled as Watson-Crick-Franklin (WC) and wobble base pairs (hereafter referred to as ‘base pairs’ or ‘pairs’). All other base pairs are defined as non-canonical base pairs and analyzed separately. Crossed base pairs (pseudoknots) were defined as non-nested canonical base pairs, that is, any canonical base pair (i,j) for which another canonical base pair (k,l) existed with i < k < j < l or k < i < l < j. Singlet base pairs were defined as any canonical base pair that was not part of a stem, that is, (i,j) such that there was no neighboring canonical base pair between i + 1 and j − 1 or between i − 1 and j + 1. Intermolecular base pairs were identified as any canonical base pair between nucleotides in different chains.
It is worth noting that DSSR is actively supported, and I always strive to respond to users’ questions via email or (preferably) on the 3DNA Forum quickly and concretely. If you have any questions about DSSR or need clarifications, please feel free to contact me. Additionally, I monitor 3DNA/DSSR citations in the literature and proactively address issues that come to my attention when necessary.
