ADMETlab 3.0: an updated comprehensive online ADMET prediction platform enhanced with broader coverage, improved performance, API functionality and decision support.
ADMETlab 3.0 is a comprehensive online platform that has been significantly updated to enhance its capabilities for predicting ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) parameters, along with physicochemical properties and medicinal chemistry characteristics crucial for drug discovery . This version is the second update to the web server, addressing limitations of its predecessors by offering broader coverage, improved performance, API functionality, and robust decision support .
Key Enhancements and Features of ADMETlab 3.0:
- Expanded Data and Endpoints: ADMETlab 3.0 incorporates 119 features, representing an increase of 31 features compared to the previous version . The updated database is 1.5 times larger, containing over 400,000 entries, which contributes to its broader coverage and improved prediction accuracy .
- Advanced Architecture and Performance: The platform utilizes a multi-task Deep Message Passing Neural Network (DMPNN) architecture, combined with molecular descriptors . This methodological advancement ensures rapid calculation speeds for each endpoint simultaneously, while also achieving superior performance in terms of accuracy and robustness .
- API Functionality: To meet the increasing demand for programmatic access to large datasets, an Application Programming Interface (API) has been introduced in ADMETlab 3.0 . This allows users to integrate ADMETlab 3.0's prediction capabilities into their own computational workflows, facilitating high-throughput analysis and automation in drug discovery .
- Uncertainty Estimates: A significant addition in this version is the inclusion of uncertainty estimates in the prediction results . This feature is invaluable for researchers, as it aids in the confident selection of candidate compounds for further studies and experiments by providing a measure of reliability for each prediction .
- Public Accessibility: ADMETlab 3.0 is freely available for public access without the need for registration. It can be accessed at https://admetlab3.scbdd.com .
- Authorship and Institutional Affiliations: The development of ADMETlab 3.0 involved researchers from multiple institutions, including the Xiangya School of Pharmaceutical Sciences, Central South University (Changsha, Hunan, P.R. China), the School of Chinese Medicine, Hong Kong Baptist University (Kowloon, Hong Kong SAR), and the School of Computer Science, National University of Defense Technology (Changsha, Hunan, P.R. China) . The corresponding authors include Li Fu, Shaohua Shi, Jiacai Yi, and Dongsheng Cao . The publication "Nucleic Acids Research," where ADMETlab 3.0 was described, has an impact factor of 13.1 .
Applications and Impact of ADMETlab 3.0 in Research:
ADMETlab 3.0 has been extensively utilized in various research studies to predict and evaluate the ADMET profiles of diverse compounds, contributing significantly to drug discovery, toxicology, and environmental health risk assessments.
-
Drug Discovery and Design:
- Anti-inflammatory and Immunomodulatory Agents: Batatasin-III, a compound from Bletilla striata, was evaluated using ADMETlab 3.0, which predicted favorable drug-likeness, bioavailability, and low toxicity, supporting its potential as a therapeutic agent for ulcerative colitis through multitarget modulation of inflammation, kinase regulation, and epithelial repair .
- Anticancer Drug Development: Novel analogues of nilutamide (NLM) were designed as potential antiandrogen agents for prostate cancer therapy . ADMETlab 3.0 was used to calculate ADMET scores for these analogues, with NLM34 and NLM40 showing optimized pharmacokinetic profiles and reduced toxicity compared to the parent compound, making them promising candidates . Similarly, new 4-benzenesulfonamide derivatives of pyrazolo[1,5-a][1,3,5]triazine were synthesized and evaluated for anticancer activity, with ADMETlab 3.0 predicting favorable ADMET profiles for compounds 4, 5, and 7, which exhibited potent activity against various cancer cell lines and low toxicity .
- Antiplasmodial Compounds: n-Hexadecanoic acid (HA) was investigated for its multi-stage antiplasmodial potency and toxicity . ADMETlab 3.0 was employed to predict its absorption, distribution, metabolism, excretion, and toxicity profiles, which, along with in vivo studies, indicated that HA is safer at lower doses and is a potential drug candidate for malaria .
- Erectile Dysfunction Treatment: The ADMET properties of molecules like L-arginine, L-citrulline, resveratrol, alpha-lipoic acid, and rutin, potentially involved in erectile dysfunction (ED) treatment, were determined using ADMETlab 3.0 . The analysis showed L-arginine and L-citrulline to have low toxicity and positive therapeutic effects, while resveratrol showed promising data but required further investigation into its potential toxicity and metabolic interactions .
- Major Depressive Disorder Treatment: A duloxetine (DLX) analog, DLX48, was designed and evaluated in silico as a potential next-generation treatment for Major Depressive Disorder (MDD) . ADMETlab 3.0 predictions indicated that DLX48 complies with drug-likeness rules, exhibits improved aqueous solubility and optimal lipophilicity, and crucially, shows no predicted hepatotoxicity or genotoxicity, suggesting enhanced pharmacological efficacy and a reduced toxicity profile .
-
Toxicology and Safety Assessment:
- Environmental Pollutants:
- 6PPDQ-Induced Hepatotoxicity: ADMETlab 3.0 was utilized to predict the physicochemical properties and multiorgan toxicity of 6PPDQ, a rubber tire-derived environmental pollutant . The platform's predictions contributed to a comprehensive mechanistic framework for 6PPDQ-induced liver injury, highlighting mechanisms such as apoptosis, inflammation, and lipid metabolic disturbances .
- Microplastics (MPs) and Allergic Rhinitis: The toxicity profiles of typical microplastics (polyethylene, polypropylene, polyvinyl chloride, polystyrene) were evaluated using ADMETlab 3.0 . The analysis revealed that MPs exhibit significant respiratory and ocular toxicity, influencing allergic rhinitis pathogenesis through pathways involving apoptosis, mitochondrial autophagy, and inflammation .
- DINCH-Induced Hepatotoxicity: ADMETlab 3.0 was used to predict the hepatotoxicity and carcinogenicity of Diisononyl cyclohexane-1,2-dicarboxylate (DINCH), a phthalate substitute . The predictions helped elucidate the primary mechanisms by which DINCH may induce hepatotoxicity, primarily by upregulating TNF, TP53, and PPARG .
- 6PPD and 6PPD-quinone Respiratory Toxicity: ADMETlab 3.0, along with ProTox-II, was used to predict the respiratory hazard potential of 6PPD and its ozone-derivative, 6PPD-quinone . The study found that these emerging pollutants disrupt mitochondrial energy metabolism, dysregulate apoptotic pathways, and activate inflammatory cascades, leading to respiratory inflammation .
- Organophosphorus Pesticides (OPs): ADMETlab 3.0 was applied to predict toxicity indicators for parent OPs and their environmental transformation products . The findings indicated that unrestricted OPs like phorate, parathion, and chlorpyrifos have a high probability of toxicity, and their transformation products pose similar comprehensive toxicity risks, leading to the creation of a "special attention list" for OPs .
- Drug-Induced Toxicity:
- Vorasidenib Safety Profile: An in silico analysis of Vorasidenib, an IDH1/2 inhibitor approved for grade 2 astrocytomas and oligodendrogliomas, was conducted using ADMETlab 3.0 and other computational tools . The results suggested potential risks of drug-induced liver injury (DILI), hepatotoxicity, neurotoxicity, nephrotoxicity, cardiotoxicity (including hERG channel blockade), genotoxicity, and carcinogenicity, underscoring the need for careful patient monitoring .
- Arecoline-Induced Oral Cancer: Toxicity profiling of arecoline, a carcinogen from betel nut, was performed using ProTox-3.0 and ADMETlab databases . The study investigated the molecular mechanisms of arecoline-induced oral cancer, identifying core targets like TP53, TNF, IL6, and CASP3, which interfere with cellular growth, inflammatory responses, and apoptotic mechanisms .
- Environmental Pollutants:
Comparison with ADMETlab 2.0:
While ADMETlab 3.0 represents a significant upgrade, ADMETlab 2.0 has also been used in various studies for ADMET predictions.
- Pharmacokinetic and Bioactivity Prediction: ADMETlab 2.0 was utilized to predict the pharmacokinetic properties and bioactivity profiles of the main constituents of Foeniculum vulgare essential oil (FVEO) . This was part of a study investigating FVEO's effects on scopolamine-induced cognitive deficits in zebrafish, where FVEO was found to improve cognitive performance and reduce oxidative stress .
- Comparison of Thiosemicarbazide and Semicarbazide Derivatives: ADMETlab 2.0 software was used for in silico analysis to compare the ADMET profiles of thiosemicarbazide and semicarbazide derivatives with proven antitumor activity . The comparative analysis revealed that semicarbazides generally have more favorable intestinal absorption, lower biological activity but higher selectivity, and lower risk of drug interactions . Thiosemicarbazides, on the other hand, showed a higher probability of metabolic activity with increased toxicity, higher plasma protein binding, a lower unbound fraction, and a longer half-life . The study concluded that semicarbazides are better candidates for anticancer drug trials due to their more favorable pharmacokinetic and pharmacodynamic profiles and lower toxicity .
In summary, ADMETlab 3.0 is a robust and highly capable platform that significantly advances in silico ADMET prediction. Its expanded coverage, improved performance, API integration, and uncertainty estimates make it an indispensable tool for accelerating drug discovery and enhancing toxicology assessments across a wide range of applications.