DPDR-CPI, a server that predicts Drug Positioning and Drug Repositioning via Chemical-Protein Interactome

Abstract

The cost of developing a new drug has increased sharply over the past years. To ensure a reasonable return-on-investment, it is useful for drug discovery researchers in both industry and academia to identify all the possible indications for early pipeline molecules. For the first time, we propose the term computational “drug candidate positioning” or “drug positioning”, to describe the above process. It is distinct from drug repositioning, which identifies new uses for existing drugs and maximizes their value. Since many therapeutic effects are mediated by unexpected drug-protein interactions, it is reasonable to analyze the chemical-protein interactome (CPI) profiles to predict indications. Here we introduce the server DPDR-CPI, which can make real-time predictions based only on the structure of the small molecule. When a user submits a molecule, the server will dock it across 611 human proteins, generating a CPI profile of features that can be used for predictions. It can suggest the likelihood of relevance of the input molecule towards ~1,000 human diseases with top predictions listed. DPDR-CPI achieved an overall AUROC of 0.78 during 10-fold cross-validations and AUROC of 0.76 for the independent validation. The server is freely accessible via http://cpi.bio-x.cn/dpdr/.

Uncovering the Pharmacological Mechanism of Stemazole in the Treatment of Neurodegenerative Diseases Based on a Network Pharmacology Approach

“…The 3D structural file or SMILES string was submitted to the PharmMapper (http://lilab-ecust. cn/pharmmapper/) [55], DRAR-CPI (http://cpi.bio-x.cn/drar/) [56], DPDR-CPI (http://cpi.bio-x.cn/ dpdr/) [57], TargetNet (http://targetnet.scbdd.com/) [58], and ChemMapper (http://lilab-ecust.cn/ chemmapper/) [59] servers to predict the targets of stemazole. Homo sapiens was chosen as the organism.…”

Molecular Docking for Prediction and Interpretation of Adverse Drug Reactions

“…Like molecular docking scores, computing structural fingerprints requires no more information than the structures. We generated seven different types of structural fingerprints used in our previous work [25] to serve as baselines for comparison of our molecular docking model. The seven structural fingerprints are E-state, Extended Connectivity Fingerprints (ECFP)-6, Functional-Class Fingerprints (FCFP)-6, Fingerprint 4 (FP4), Klekota-Roth method, Molecular ACCess System (MACCS) and PubChem structural descriptors (labeled E-state, ECFP6, FCFP6, FP4, KR, MACCS and PubChem, respectively).…”

“…With recent increases in computing power, machine learning has played an important role in healthcare data analysis and prediction. In previous studies, we utilized machine learning models to predict drug indications [25] and drug-drug interactions [17] based on the drug-target binding profiles generated from molecular docking. We showed that the chemical-protein interactome (CPI), a collection of drugtarget binding profiles simulated with molecular docking, can be used to both predict drug pharmacologic actions and to provide possible biologic explanations for them.…”

Online structure-based screening of purchasable approved drugs and natural compounds: retrospective examples of drug repositioning on cancer targets

“…Several webservers are available for target fishing [ 12 ], i.e. to try to identify a protein target for a given compound, and thus can be applied to drug repositioning [ 13 – 15 ]. Yet, to the best of our knowledge, the only webserver that provides a prepared virtual library of approved drugs (1852 molecules approved by the FDA between 1939 and 2017) and a facility to screen these compounds over the users' selected protein target is the e-Drugs3D webserver [ 16 ].…”