High-throughput solubility measurements, experimental

As a key first step towards oral absorption, considerable effort has been directed towards the development of computational solubility prediction [26-30]. However, partly due to a lack of large experimental datasets measured under identical conditions, today s methods are not sufficiently robust for reliable predictions [31]. Nonetheless, further fine-tuning of these models can be expected since high-throughput data have become available for their construction. 

Another common approach to simultaneously optimizing multiple compound properties is to apply a sequence of filters, based on criteria for experimentally measured or calculated properties. For example, the output of a high-throughput screen may be filtered to remove compounds with less than 80% inhibition of the target at 10 pM, the remaining compounds may then be filtered to remove compounds with calculated logP >5, predicted solubility <10 pM, predicted human intestinal absorption <30%, and possibly many other criteria. The hope is that one or more compounds will pass all of these criteria. 

---