Present drug-likeness filters in medicinal chemistry during the hit and lead optimization process: how far can they be simplified?

During the past decade, decreasing the attrition rate of drug development candidates reaching the market has become one of the major challenges in pharmaceutical research and drug development (R&D). To facilitate the decision-making process, and to increase the probability of rapidly finding and developing high-quality compounds, a variety of multiparametric guidelines, also known as rules and ligand efficiency (LE) metrics, have been developed. However, what are the 'best' descriptors and how far can we simplify these drug-likeness prediction tools in terms of the numerous, complex properties that they relate to?