Alzheimer's Disease (AD) is a universal neurodegenerative disease with the feature of progressive dementia. It begins with mild memory loss and, as the disease progresses with age, leads to the degeneration of specific nerve cells involved in memory, language, and cognition. The disease causes great suffering to patients and their families. Currently, there are only seven FDA-approved drugs for the treatment of AD, which can only temporarily slow down the deterioration of symptoms and cannot reverse the disease process. Screening for inhibitors that can interact with proteins associated with AD plays a key role in the development of AD drugs. Huge amounts of such inhibitors are scattered in numerous published articles, which is inconvenient for researchers to explore the drug candidates for AD. IPAD-DB is the first web-accessible database for experimentally verified inhibitors of proteins associated with Alzheimer's Disease. We hope that IPAD-DB can play a role in the screening of AD drug candidates and the investigation of the pathogenesis of AD.
Inhibitors curated in the database include natural compounds, synthetic compounds, drugs, natural extracts, and nano-inhibitors. To date, the database has compiled 4,804 entries, each representing a correspondently relationship between an inhibitor and its target protein. The IPAD-DB database is not finished completely now, and more records will be updated regularly after manual curation. For the five types of inhibitors, we provide abundant annotations through compilation and integration, using a wide range of public resources. This annotation covers six aspects, including (i) basic information, (ii) biological activity data, (iii) physicochemical properties, (iv) ADMET properties, (v) pharmacokinetic prediction, and (vi) Druglikeness prediction. Inhibitor names and structures are first retrieved from collected literature and searched in the PubChem database to gather basic information about the compound, including name, molecular formula, molecular weight, structure (2D and 3D), IUPAC name, InChI, InChIKey, and canonical SMILES. The collected inhibitors are cross-linked to existing databases for the further study. In instances where molecules could not be retrieved from PubChem, we utilized StoneMIND Collector software (StoneMIND Release 2021–9. StoneMIND Collector, Beijing StoneWise Technology Co Ltd, CN, 2021) to map the compound's 2D structure based on the provided information in the literature. These structures were then converted to IUPAC Name, InChI, InChIKey, and Canonical SMILES formats and stored in IPAD-DB. Biological activity data (including Ki, EC50, IC50, etc.), toxicity, reactive oxygen species (ROS), metal chelation, blood-brain barrier permeability (BBB), target protein, effect, and literature URL are extracted from the collected literature. Two online websites SwissADME, and ADMETlab 2.0 were used to calculate the physicochemical properties, ADMET properties, pharmacokinetic prediction, and druglikeness of inhibitors. For natural extracts and nano-inhibitors are mixtures and have no single structure, so some physical and chemical properties can not be predicted and provided. All datasets and annotations are freely accessible to all users.
Article is ongoing.