I am encountering issues while preparing a PDB file of a substrate (UDP-MurNAc-L-Ala, with added charges and hydrogens) for molecular docking with AutoDock Vina. The conversion to the PDBQT format results in Vina failing to recognize the ligand due to the presence of non-standard atom types, represented as "UNK" in the HETATM records (e.g., HETATM 112 C UNK 0 100.803 30.414 35.420 0.00 0.00 C).

Manually editing these "UNK" labels to their expected element type (e.g., "C") allows Vina to read the file, but the resulting 3D structure of the ligand is severely stretched and geometrically incorrect. Although all coordinates stay same.

What is the recommended workflow or specific tools to correctly resolve these abnormal atom type assignments in my PDB ligand file and ensure the preservation of its accurate three-dimensional structure for subsequent docking studies using AutoDock Vina?

Clipped files are original .pdb and corrupted one respectively