Is the Problem Tractable within a Single Program

A reliable druglikeness predictor would give high prediction scores only to compounds that have satisfactory properties based on aU of these criteria, or in other words, there would be few false positives. There has been considerable effort expended over the last 10 to 20 years in modeling individual components of this process, including solubility [18-36], ADME properties [53-70], and toxicities [71-88]. Individually, each of tliese predictions has false positives and false negatives, so it is difficult to expect 

Do We Have a Training Set that Will Allow Us to Address the Issue.  

Most prediction tools for druglikeness use the MDL Available Chemicals Directory (ACD) [89] as a list of nondrugs and one of three databases as a list of drugs MDL Drug Data Report (MDDR) [89], MDL Comprehensive Medicinal Chemistry (CMC) [89], or Derwent World Drug Index (WDI) [90]. 

MDDR is a database derived principally from the patent literature, journals, and meetings. It contains the structures of approximately 130000 compounds of which approximately 1000 are launched drugs. WDI is a Derwent database of approximately 73 000 marketed drugs and development compounds, drawn from journals, scientific meetings, and approved name lists. CMC is a list of about 8400 compounds taken from the Drug Compendium in Pergamon s Comprehensive Medicinal Chemistry [94] and also from the United States Approved Names list. 

---