DNA Scanning module
Submodules
Module contents
HADDOCK3 module for DNA base-pair scan.
This module performs a mutagenesis scan of DNA base pairs for the model(s) generated in the previous step of the workflow. DNA is double stranded and its bases are engaged in Watson-Crick base pairs (A:T and G:C). Mutating a single base in isolation would break the base pair and is not physically meaningful, so every mutation performed by this module is a double mutation: for each selected interface nucleotide, the nucleotide is mutated into a target base and its base-pairing partner on the complementary strand is simultaneously mutated into the complementary base, so that a valid Watson-Crick pair is preserved (e.g. A:T -> G:C). The differences in the haddock score and its individual components between the wild type and each mutant base pair are calculated, thus providing a measure of the impact of such mutations. By default, all four possible bases are tested for each selected interface nucleotide (the wild-type base is skipped). Such difference (delta_score) is always calculated as:
delta_score = score_wildtype - score_mutant
As the score are typically negative with lower values being better, a _positive_ delta_score indicates that the mutation is stabilizing while a _negative_ delta_score indicates that the mutation is destabilizing.
If cluster information is available, the module will also calculate the average haddock score difference for each cluster of models. For each mutation, a Z score is calculated as:
Z = (delta_score - mean) / std
where mean and std are the mean and standard deviation of the delta_score over all the mutations.
The module will also generate plots of the DNA scan data, showing the distribution of the delta_score (and every component) for each base-pair mutation at the interface.
You can use the parameters below to customize the behavior of the module:
scan_bases: list of target bases to be tested for each nucleotide. By default all four DNA bases (DA, DC, DG, DT) are tested.
bp_cutoff: distance cutoff (Å) used to detect Watson-Crick base pairs from the distance between the hydrogen-bonding ring nitrogens.
chains: list of chains to be considered for the DNA scan. In some cases you may want to limit the analysis to a single chain.
output_mutants: if True, the module will output the models with the mutations applied (only possible if there is only one model)
output_bfactor: if True, the module will output the non-mutated models with the rescaled delta_score in the B-factor column
plot: if True, the module will generate plots of the DNA scan data
splitplot: if True, the scan plot shows one panel per energy component; if False (default) all components are overlaid in a single panel
resdic: list of residues to be used for the scanning. An example is:
>>> resdic_A = [1,2,3,4] >>> resdic_B = [2,3,4]
Only nucleic acid (DNA) residues at the interface are mutated; protein and other residues are ignored. Interface nucleotides for which no base-pairing partner can be found are skipped, since a double (base-pair) mutation cannot be constructed for them.
When applying a mutation, the deoxyribose-phosphate backbone is always kept for both nucleotides of the pair. For mutations between bases of the same ring type (purine <-> purine, i.e. DA <-> DG, or pyrimidine <-> pyrimidine, i.e. DC <-> DT), the shared base ring atoms are also kept so that the base orientation is preserved and CNS only rebuilds the differing substituents; the whole base pair is then scored in a single CNS call. For cross-type mutations, the two nucleotides swap ring type (one purine -> pyrimidine and its partner pyrimidine -> purine). Rebuilding a whole purine ring from its three anchor atoms in the same pass as the partner mutation is unreliable, so such a base pair is mutated in two sequential, CNS-regularised steps: first the purine -> pyrimidine mutation is applied and energy-minimised by CNS, then the pyrimidine -> purine mutation is applied to that minimised intermediate and scored by CNS. The score and energies of the second (final) CNS call are the ones reported and plotted. In both cases the glycosidic-region anchor atoms are kept and, for cross-type mutations, renamed to their counterpart in the target ring system (pyrimidine N1/C2/C6 <-> purine N9/C4/C8).
To keep the delta_score comparison consistent, each mutant is compared against a wild-type baseline that went through the same number of CNS minimisation passes: a one-pass baseline for same-ring-type mutants and a two-pass baseline (the wild type minimised and then re-scored) for cross-ring-type mutants.
- class haddock.modules.analysis.dnascan.HaddockModule(order, path, *ignore, init_params=PosixPath('/opt/hostedtoolcache/Python/3.10.22/x64/lib/python3.10/site-packages/haddock/modules/analysis/dnascan/defaults.yaml'), **everything)[source]
Bases:
BaseHaddockModuleHADDOCK3 module for DNA base-pair scan.