This QM/MM implementation evolved from research on organometallic complexes and catalytic systems. Since these systems are generally under 1000 atoms in size, the program had not been optimized to handle large macromolecular systems such as enzymes. In particular, the non-bonded code was not efficient for very large systems. It has been adjusted in the ADF2002.01 release, in order for the code to work efficiently on both small and large systems. Moreover, the calculation of the MM forces has been parallellized to further increase the efficiency and applicability. With the current release, systems with up to several thousand atoms have been tested without problems. Although the input is not designed to handle the topology of macromolecules such as amino acids and peptides, it can be constructed quite easily. For the large systems this amounts however to some 30000 lines; therefore, tools (pdb2adf) have been developed for creating input files automatically for a given PDB-file. This tool is available through a link on the 'Contributed Software' part of the SCM web site http://www.scm.com, but is now also part of the official release starting with adf2005.01.
Currently, only the AMBER95 and SYBYL force fields are included. This might also limit the applicability. However, the force field parameters and potentials are fairly customizable and other force fields are easily added.