Protein-glycan docking with HADDOCK3

This repository contains all data and analysis necessary to reproduce the results reported in this paper.

The repository contains the following directories

• analysis, with all the analysis scripts to generate the figures present in the paper

  • extract_glycan_stats.py: creates the files data/df_ss_all_data.tsv and data/df_clt_all_data.tsv
  • plot_SR_*.py: these scripts read files data/df_ss_all_data.tsv and data/df_clt_all_data.tsv and plot the success rates, based on the IL-RMSD, for the different docking protocols. The plots generated are analysis/figures/SR_*.png
  • estimate_refinement_impact.py and estimate_refinement_impact_max_improvement.py: they read data/aggregated_data.tsv and calculate the mean and max improvement obtained after [flexref]
  • violinplot_rmsd.py: reads data/RMSD_glycam_to_bound.txt and plots the all-atoms RMSD of the glycans conformations (generated with GLYCAM) with respect to their bound forms (analysis/figures/glycan-rmsd-violin.png)
  • plot-mdref-*.py: read data/mdref-*.tsv and create the boxplots analysis/figures/mdref-*.png, for short (SL-SB), long linear (LL) and long branched (LB) glycans
  • plot-scatter-*.py: read data/*-scatter.tsv and create the scatter plots analysis/figures/scatter-*.png, for short (SL-SB), long linear (LL) and long branched (LB) glycans

• data, with all the data generated in this study

  • df_ss_all_data.tsv: for each pdb and each docking scenario, it is reported whether there is (1) or there is not (0) at least one high, medium, acceptable, or near-acceptable model among the top 1, 5, 10, 50, 100 and 200 models. This is calculated after [rigidbody] sampling (2_caprieval) and on the rigidbody clustered models (5_caprieval) for bound docking. For unbound docking, this is calculated after rigidbody sampling (02_caprieval), on the rigidbody clustered models (06_caprieval), after [flexref] (08_caprieval), and on the flexref clustered models (11_caprieval).
  • df_clt_all_data.tsv: for each pdb and each docking scenario, it is reported whether there is (1) or there is not (0) at least one high, medium, acceptable, or near-acceptable model among the best 4 scoring models contained in the top 1, 2, 3, 4, 5, 10, 20, 30, 40 and 50 clusters. This is calculated on the rigidbody clustered models (5_caprieval) for bound docking. For unbound docking, this is calculated on the rigidbody clustered models (06_caprieval) and on the flexref clustered models (11_caprieval).

• pdb_files. This directory contains subdirectories for each PDB (relative to the protein-glycan complex). In each pdb_files/PDB/ subdirectory the following files are available:

  • 154L_r_b.pdb: protein bound structure
  • 154L_l_b.pdb: glycan bound structure
  • 154L_r_u.pdb: protein unbound structure
  • 154L_l_u.pdb: glycan unbound structure
  • 154L_l_ensemble.pdb: glycan ensemble, obtained with [mdref]
  • 154L_analysis.pdb: reference structure
  • 154L_ti-aa.tbl: AIRs file, ti-aa restraints (true interface on both protein and glycan)
  • 154L_tip-ap.tbl: AIRs file, tip-ap restraints (true-interface-protein, glycan fully passive)

• cfg_files. It contains subdirectories for each PDB (relative to the protein-glycan complex). In each cfg_files/PDB/ subdirectory the following files are available:

  • bound_default_ti-aa.toml: bound docking, default haddock scoring function, ti-aa AIRs
  • bound_vdw_ti-aa.toml: bound docking, vdW haddock scoring function, ti-aa AIRs
  • bound_vdw_tip-ap.toml: bound docking, vdW haddock scoring function, tip-ap AIRs
  • unbound_vdw_tip-ap_clust.toml: unbound docking, vdW haddock scoring function, tip-ap AIRs
  • unbound_ens_vdw_tip-ap_clust.toml: ensemble docking, vdW haddock scoring function, tip-ap AIRs
  • mdref-glycans-sf400x16.toml: sampling of glycan conformations using mdref