In some protein crystallography studies, the protein crystals are grown with the substrate/ligand/inhibitor added, thus the resulting structure contains both the ligand and the protein atom coordinates. This enables studying the protein ligand complex to understand binding, specificity, mechanism of catalysis...
On the other hand in some experiments protein crystals are grown without the ligand, thus the resulting structure contains coordinates just for the protein atoms.
On the third hand, the lucky one, your protein, overexpressed in its natural host [sometimes even in different one], co-purify with its own natural ligand, so the resulting structure contains coordinates for both the protein atoms and the sought for ligand.
To understand ligand binding the protein structure without the ligand should be compared to the structure of the protein-ligand complex.
Many published papers report both types of crystals.
For example the crystal structure of pectin methylesterase in complex with hexasaccharide VI: 2NTP is from the publication https://doi.org/10.1038/sj.emboj.7601816 which is the source of six other structures. These structures include the enzyme with different substrates, as well as the active site mutants (D178A) all of which were prepared to understand the molecular basis of substrate specificity. The conclusion after studying all of the mentioned structures is nicely summed in the sentence from the abstract:
The preferential binding of methylated sugar residues upstream of the catalytic site, and demethylated residues downstream, drives the enzyme along the pectin molecule and accounts for the sequential pattern of demethylation produced by both bacterial and plant PMEs.