GPCRMap Training, Test, and Validation Set with Results

GPCRMap_data/ contains FTMap hotspot mapping results for all GPCR structures across two probe sets and four cavity energy contribution thresholds and scripts used for metric calculation. The top-level directories are organised by probe set and energy threshold pair: FTMap_08_06/ and FTMap_04_02/ contain results generated with the original FTMap probe set at cavity energy contributions of −0.8/−0.6 and −0.4/−0.2, respectively; GPCR_Select_08_06/ and GPCR_Select_04_02/ contain the equivalent runs using the GPCR-Select probe set. Within each top-level directory, one subdirectory per structure (named {receptor}_{pdbid}{chains}) holds the raw input PDB and two Atlas result folders: Atlas_08/ or Atlas_04/ (the higher energy threshold) and Atlas_06/ or Atlas_02/ (the lower energy threshold), each containing the consensus site coordinates as individual SDF files, the mapped protein as a PDB, and a combined atlas PDB used for downstream analysis.GPCRMap_data/Scripts/ contains the processing and analysis pipeline used to generate and evaluate the mapping results above:-- Structure_preparation.sh is the upstream preparation step: it takes aligned Maptor .mae structures, runs Schrödinger's PrepWizard (disulfide assignment, protonation at pH 7.4, water sampling), extracts the requested chain(s) and reference ligand(s) via extract_chain.py, strips all other ligands, and exports a receptor-only .pdb suitable for submission to FTMap. -- get_atlas_overlap.py is the core scoring script: given a combined atlas PDB (--atlas_pdb) and a reference ligand (--lig_pdb), it uses PyMOL to compute, for the whole map and for each consensus cluster, the frag_hs (percentage of ligand heavy atoms within 2 Å of a hotspot) and hs_frag (percentage of hotspot atoms within 2 Å of the ligand) overlap metrics, writing them to a per-structure CSV. -- atlas_pdb_parser.sh is the batch driver that walks every {FTMap,GPCR_Select}_* structure and both Atlas_* thresholds, pairs each *_atlas.pdb with its reference structure in SDF_reference/, invokes get_atlas_overlap.py, and concatenates the per-structure results into a single all_ligands_combined.csv. -- SDF_reference/ holds those reference ligands as one .sdf file per structure (named {receptor}_{pdbid}{chains}.sdf), extracted from the experimental complexes and used as the ground-truth binding pose. -- Assessment_FTMapResults.py performs the downstream statistical analysis: it reads the combined coverage tables from the supplementary Excel workbook, parses the frag_hs/hs_frag strings into per-map and per-cluster values, then it computes mean coverage and hotspot-count summaries for each probe-set/energy-threshold condition, exporting them as supplementary tables.