Hongping Zhao

Aqueous solubility is an important physicochemical property of compounds in anti-cancer drug discovery. Artificial intelligence solubility prediction tools have scored impressive performances by employing regression, machine learning, and deep learning methods. The reported performances vary significantly partly because of the different datasets used. Solubility prediction on novel compounds needs to be improved, which may be achieved by going deeper with deep learning. We constructed deeper-net models of ~20-layer modified ResNet convolutional neural network architecture, which were trained and tested with 9,943 compounds encoded by molecular fingerprints. Retrospectively tested by 62 recently-published novel compounds, one deeper-net model outperformed four established tools, shallow-net models, and four human experts. Deeper-net models also outperformed others in predicting the solubility values of a series of novel compounds newly-synthesized for anti-cancer drug discovery. Solubility prediction may be improved by going deeper with deep learning. Our deeper-net models are accessible at http://www.npbdb.net/solubility/index.jsp.

Cardiovascular drug research and development (R&D) has been in active state and continuously attracts attention from the pharmaceutical industry. However, only one individual drug can eventually reach the market from about the 10,000 compounds tested. It would be useful to learn from these failures when developing better strategies for the future. Discontinued drugs were identified from a search performed by Thomson Reuters Integrity. Additional information was sought through PubMed, ClinicalTrials.gov, and pharmaceutical companies search. Twelve compounds discontinued for cardiovascular disease treatment after reaching Phase I–III clinical trials from 2016 to 2018 are detailed in this manuscript, and the reasons for these failures are reported. Of these, six candidates (MDCO-216, TRV027, ubenimex, sodium nitrite, losmapimod, and bococizumab) were dropped for lack of clinical efficacy, the other six for strategic or unspecified reasons. In total, three candidates were discontinued in Phase I trials, six in Phase II, and three in Phase III. It was reported that the success rate of drug R&D utilizing selection biomarkers is higher. Four candidate developments (OPC-108459, ONO-4232, GSK-2798745, and TAK-536TCH) were run without biomarkers, which could be used as surrogate endpoints in the 12 cardiovascular drugs discontinued from 2016 to 2018. This review will be useful for those involved in the field of drug discovery and development, and for those interested in the treatment of cardiovascular disease.

Introduction: About 10,000 compounds will be tested for an individual drug to eventually reach the market. It might be helpful recapitulating previous failures and identifying the main factors of the disappointments. Areas covered: In this review, the author(s) detailed the 7 cardiovascular compounds discontinued after reaching animal studies or Phase I—III clinical trials during 2015. Meanwhile, the reasons for these discontinuations were reported. Among these drugs, most discontinuations (6 drugs) were attributed to lack of efficacy. In general, failures due to lack of efficacy and safety demonstrate the need for the development of more predictive animal models. However, recent related studies showed that the absence of toxicity in animals provided little or virtually no evidential weight that adverse drug reactions would also be absent in humans. In this case, microdosing and collaborating more closely with biotech companies may be the better choices to improve the success ratio. Expert opinion: Future researches may benefit from the seven developments and investigators conducting similar studies may learn from these failures.