Table 1. Description of column names in the Twilight script.

This table contains a reference sheet for all columns shown in the Twilight main table, supplying remarks on their origin. If a datum is unavailable for a particular PDB entry, the description mentions its default value for the corresponding column.

Column name	Description
Rank†	Initial sort order upon loading, which is a combination of descending score S and grouping by PDB four-character code.
PDBID‡	RCSB Protein Data Bank (PDB) four-character accession code.
LigNm‡	Three-letter PDB ligand name.
ResNr‡	Unique residue identifier. This has the general format CNNNNI, where C is the chain name, NNNN is the four-digit residue number and I is the insertion code. The residue number NNNN is right-justified and the insertion code I is mostly the space character. Residue number 66 of chain A is therefore represented as ‘A  66 ’. Notice that the RSCC values are not reported by the EDS separately when alternate conformers are present and therefore the alternate conformation indicators are not included.
RSCC§	Real-space correlation coefficient for the ligand retrieved from EDS. Used in the computation of the score S.
OWAB§	Occupancy-weighted average B factor of the ligand, also retrieved from EDS.
MolWt‡	Molecular weight of the ligand in units of Da.
Charge¶	Ligand charge, computed from its chemical formula.
#Heavy¶	Number of non-H atoms that the ligand is composed of.
#Hydrog¶	Number of H atoms for the ligand.
#Atoms¶	The overall number of atoms; that is, #Heavy + #Hydrog.
DepDate‡	Protein structure deposition date according to the HEADER record from the PDB.
Resol‡	Experimental resolution of the crystal structure in units of Å. Used in the calculation of the score S.
Rwork‡	Working-set R value. A negative number indicates its absence.
Rfree‡	Test-set R value (R free). A negative number indicates its absence.
Robs‡	Observed-set (working set and test set taken together) R value. A negative number indicates its absence.
Softwre†	Software used for crystallographic refinement as stated in REMARK 3 of the PDB entry. In the case that multiple refinement programs are stated, this becomes a comma-separated list of program names.
Jrnl‡	Abbreviated journal name if an article describing the structure has been published. Contains ‘To be Published’ if no publication is associated with the PDB entry.
PMID‡	NCBI PubMed identifier of the primary citation, if the structure has been published; otherwise, the cell is empty.
Class†	Ligand type. Frequently encountered crystallization buffer molecules are conveniently flagged with the letter ‘B’, otherwise the hyphen character ‘-’ is printed.
Valid†	This column contains the letter ‘Y’ (RSCC ≤ 0.6) or ‘G’ (RSCC ≥ 0.95) to indicate that the ligand has passed all of the filter criteria described in §2. The letter ‘A’ is used for heterogroups from a manually imported structure.
Score†	Combination of the real-space correlation coefficient and the experimental resolution into a single score S, which is used for ranking the list of ligands.
Comment	Author comment if the particular ligand has been analyzed; otherwise, it is empty. We have blanked all our in-house-generated annotations except for false positives. However, users are especially encouraged to add their own annotations in this column.

^†Value resulting from the filtering procedure mentioned in §2.

^‡Value directly taken from the result list returned by the PDB Advanced Search Interface query described in §2.

^§Value retrieved from the EDS server.

^¶Value derived from the chemical formula returned by the PDB Advanced Search Interface query described in §2.