Recently, I read the preprint of Gordan et al. (2025), titled "High-throughput characterization of transcription factors that modulate UVdamage formation and repair at single-nucleotide resolution". In the METHODS section on "Structural analysis of AlphaFold 3 predicted TF-DNA complexes", the authors introduced two geometric parameters to characterize dipyrimidines, as detailed below:

Base-step d22 distance, d64 distance, of dipyrimidines were computed for each base-step per DNA strand using custom PyMOL python scripts⁷⁴. d22 was defined as the distance in Ångstroms (Å) between the C5-C6 bond midpoints between adjacent pyrimidines. d64 was defined as the Å distance between the 5' pyrimidine's C5 and X4 (either O or N) attached to the 3' pyrimidine's C4.

These d22 and d64 parameters are well-defined and straightforward to calculate (see the figure below for illustrative examples). They can be integrated seamlessly with DSSR's infrastructure, requiring minimal additional coding effort. As a result, I have decided to implement them into DSSR.

For example, in the case of the MyoD bHLH domain-DNA complex (PDB ID: 1mdy), running the following DSSR (v2.7.1) command:

x3dna-dssr -i=1mdy.pdb --json -o=1mdy.json

generates a JSON file (1mdy.json), which contains the following information under the nts section for E.DT1 (connected with E.DC2): "d22": 4.014 and "d64": 3.655."

The default human-readable output file, dssr-torsions.txt, now includes two additional columns for d22 and d64 under the section titled 'Main chain conformational parameters,' as shown below.

          nt        d22     d64
 1     T E.DT1     4.01    3.66
 2     C E.DC2      ---     ---
 3     A E.DA3      ---     ---
 4     A E.DA4      ---     ---
 5     C E.DC5      ---     ---
 6     A E.DA6      ---     ---
 7     G E.DG7      ---     ---
 8     C E.DC8     4.13    4.44
 9     T E.DT9      ---     ---

Note that pseudouridine (Ψ) is excluded from the calculation of d22 and d64 parameters of a dipyrimidine base step.

The implementation of d22 and d64 parameters in DSSR is a clear example of my proactive approach to enhancing the software's functionality. Users are always encouraged to reach out with requests for new features or improvements, as well as to report any bugs or ask questions.

References

1. Gordan R, Wasserman H, Chi B, Bohm K, Duan M, Sahay H, et al. High-throughput characterization of transcription factors that modulate UVdamage formation and repair at single-nucleotide resolution. 2025. https://doi.org/10.21203/rs.3.rs-8197218/v1.