The efficiency of quantum mechanical (QM) computational tools has increased remarkably over the last decade. Accurate QM calculations are no longer limited to small isolated molecular systems with tens or at most a few hundred atoms. Semi-empirical QM (SQM) methods, such as PM6-D3H4X, and hybrid QM/molecular mechanics (MM) methods [1], where only a small part of the system is treated at the QM level, can be applied to real chemical problems of practical importance. In addition to higher accuracy, QM methods are free from any empiricism (except for system-independent parameters in the underlying computational methods) and are not tuned to a specific goal, which makes them universally applicable. The same is true of the recently emerging general-purpose machine learning (ML) potentials.[2]

An important area where the interest in QM calculations is now growing rapidly is computer-aided drug design [3,4]. Efficient QM or ML methods, applicable to large systems such as protein-ligand complexes, bring the much needed accuracy that the more approximate methods commonly used in this field cannot provide. The cornerstone of structure-based computer-aided drug design (CADD) is the calculation of protein-ligand binding affinities, and various QM-based approaches have already demonstrated excellent performance there [5]. It is clear that the use of QM calculations, which are based on sound physical principles that are general across different protein targets and different ligand chemistries, offers unique advantages. QM methods also play an important supporting role in the calculation of molecular properties used in other types of calculations.

However, applications of QM methods in CADD are still far from mainstream, and much work needs to be done to make them more widely applicable. However, the potential benefits are well worth the effort. The proposed workshop is designed to facilitate these developments.