Initiating heavy-atom-based phasing by multi-dimensional molecular replacement

A strategy is presented to set up an n-dimensional molecular-replacement parameter matrix (MRPM) search using anomalous difference Fourier maps from related data sets to uncover weak, but correct, molecular-replacement solutions for heavy-atom substructure determination and subsequent experimental phasing.


Supplementary
. Generation of the search model library. Generation of the library is divided into 4 steps.
Step 1: Identify the domains to use.
Step 2: Identify the scaffolds used to place the domains into representing different possible conformations and place the domains into these scaffolds.
Step 3: Identify and generate a number of truncations removing different domains, since one incorrectly placed domain can make the difference between success and failure.
Step 4: Prune the atoms of the models to generate variations. In this particular case only two different pruning schemes were used: all atom or reduction to poly-alanine. Figure S2. Superposition of the 15 starting models after step 2 in Supp. Figure 1. All models are superposed on the P domain and each model has a distinct color. The conformational variation obtained by using different scaffolds is evident. The functional cycles of P-type ATPases is characterized by four principal conformations (Møller et al., 2010). The colors are chosen to emphasize that the models fall into these classes: Red shades are the Occluded outwards facing forms. Blue shades are the occluded transition state forms. Green shades are the open outward facing forms. Yellow shades are the inward facing forms.

Supplementary Table S7. r.m.s.d. of search-models with no N domain (A+P+TM16).
Pdbs are numbered 1-15. Their order is identical to the order seen in Figure 2C and Supplementary Table 4, with the addition of the H+-ATPase (3b8c) added as a scaffold (pdb15). Red notes r.m.s.d. below 1 Å.

Supplementary Table S8. r.m.s.d. of search-models with no A domain (N+P+TM16).
Pdbs are numbered 1-15. Their order is identical to the order seen in Figure 2C and Supplementary  Figure 2C and Supplementary Table S4, with the addition of the H+-ATPase (3b8c) added as a scaffold (pdb15). Red notes r.m.s.d. below 1 Å. Note that without the transmembrane domain present the rmsd drops in may cases (compare Supplementary Table S9 to Supplementary Tables S6-S8). Especially pdb2, pdb6 and pdb7 are similar, and pdb3, pdb9 and pdb12 are similar as expected from the conformations that they represent.

Supplementary scripts
Contains 6  HA Dataset should contain the following columns: H,K,L,DANO,SIGDANO. 11) In $MRPM/models, create subdirectories called scaffold1, scaffold2 etc. One for each scaffold to test. Copy the pdb's to use as domains and scaffold into each scaffold-subdirectory. 12) Using pymol or similar, overlay the domains to the scaffold and save the final result as searchmodel.pdb. 12) Edit and run $MRPM/models/create_searchmodel.sh to generate the search-model library. 13) Edit $MRPM/input/phaser_and_analysis_for_setup.sh to set up the parameters to scan in the individual MR runs. 14) Edit $MRPM/input/evaluate_for_setup.sh to set up the parameters used to calculate the anomalous difference maps. 15) Edit and run $MRPM/setup_search.sh to set up the directory structure and input files for the search. 16) Edit and run $MRPM/start_runs_setup.sh to set up a 'start_runs.sh' file that will initiate MRPM on a given number of cores. 17) Run $MRPM/start_runs.sh to initiate MRPM. 18) During and after the runs have finished run $MRPM/eval_result.sh to list the results from individual runs that have completed.
Use GNUPLOT or similar to plot the results if desired.