Application scenario-oriented molecule generation platform developed for drug discovery.

Design of molecules for candidate compound selection is one of the central challenges in drug discovery due to the complexity of chemical space and requirement of multi-parameter optimization. Here we present an application scenario-oriented platform (ID4Idea) for molecule generation in different scenarios of drug discovery. This platform utilizes both library or rule based and generative based algorithms (VAE, RNN, GAN, etc.), in combination with various AI learning types (pre-training, transfer learning, reinforcement learning, active learning, etc.) and input representations (1D SMILES, 2D graph, 3D shape, binding site, pharmacophore, etc.), to enable customized solutions for a given molecular design scenario. Besides the usual generation followed screening protocol, goal-directed molecule generation can also be conducted towards predefined goals, enhancing the efficiency of hit identification, lead finding, and lead optimization. We demonstrate the effectiveness of ID4Idea platform through case studies, showcasing customized solutions for different design tasks using various input information, such as binding pockets, pharmacophores, and compound representations. In addition, remaining challenges are discussed to unlock the full potential of AI models in drug discovery and pave the way for the development of novel therapeutics.

Network pharmacology-based approach to investigate the mechanisms of Zingiber officinale Roscoe in the treatment of neurodegenerative diseases.

Neurodegenerative diseases (NDDs) are chronic neurological disorders associated with cognitive or motor dysfunction. As a common spice, Zingiber officinale Roscoe has been used as a medicine to treat a variety of NDDs. However, at the molecular level, the mechanisms of Z. officinale in treating of NDDs have not been deeply investigated. In this study, network pharmacology method, molecular docking, Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were used to predict the mechanisms of Z. officinale in the treatment of NDDs. After a series of biological information analyses, five core targets were obtained, including heme oxygenase 1 (HMOX1), acetylcholinesterase (AChE), nitric oxide synthase (NOS), catechol-O-methyl-transferase (COMT), and metabotropic glutamate receptor 5 (mGluR5). Compounds 75, 68, 46, 67, 69, 49, 66, 50, 34, and 64 were identified as the main components of Z. officinale in the treatment of NDDs. The crucial pathways mainly include neuroactive ligand-receptor signaling pathways, cyclic adenosine monophosphate signaling pathways, dopamine synaptic signaling pathways, and so on. Besides, in vitro experiments by AChE inhibitory activities assay and neuroprotective activities against H2 O2 -induced injury in human neuroblastoma SH-SY5Y cells validated the reliability of the results of network analysis. PRACTICAL APPLICATIONS: Zingiber officinale Roscoe is widely used as a traditional spice and herbal medicine. It contains a number of active ingredients, which have shown activities on anti-neurodegenerative diseases (NDDs). In this paper, the potential mechanism of Z. officinale in the treatment of NDDs is explored through network pharmacology, and it was verified by in vitro experiments. The mechanism was not only clarified at the system level but also proved to be effective at the biological level. The results can be used as a reference for Z. officinale in the treating of NDDs.

Mechanism of Side Chain-Controlled Proton Conductivity in Bioinspired Peptidic Nanostructures.

Bioinspired peptide assemblies are promising candidates for use as proton-conducting materials in electrochemical devices and other advanced technologies. Progress toward applications requires establishing foundational structure-function relationships for transport in these materials. This experimental-theoretical study sheds light on how the molecular structure and proton conduction are linked in three synthetic cyclic peptide nanotube assemblies that comprise the three canonical basic amino acids (lysine, arginine, and histidine). Experiments find an order of magnitude higher proton conductivity for lysine-containing peptide assemblies compared to histidine and arginine containing assemblies. The simulations indicate that, upon peptide assembly, the basic amino acid side chains are close enough to enable direct proton transfer. The proton transfer kinetics is determined in the simulations to be governed by the structure and flexibility of the side chains. Together, experiments and theory indicate that the proton mobility is the main determinant of proton conductivity, critical for the performance of peptide-based devices.

Machine learning methods to predict the cultivation age of Panacis Quinquefolii Radix.

BACKGROUND: American ginseng (AG) is a valuable medicine widely consumed as a herbal remedy throughout the world. Huge price difference among AG with different growth years leads to intentional adulteration for higher profits. Thus, developing reliable approaches to authenticate the cultivation ages of AG products is of great use in preventing age falsification. METHODS: A total of 106 batches of AG samples along with their 9 physicochemical features were collected and measured from experiments, which was then split into a training set and two test sets (test set 1 and 2) according to the cultivation regions. Principle component analysis (PCA) was carried out to examine the distribution of the three data sets. Four machine learning (ML) algorithms, namely elastic net, k-nearest neighbors, support vector machine and multi-layer perception (MLP) were employed to construct predictive models using the features as inputs and their growth years as outputs. In addition, a similarity-based applicability domain (AD) was defined for these models to ensure the reliability of the predictive results for AG samples produced in different regions. RESULTS: A positive correlation was observed between the several features and the growth years. PCA revealed diverse distributions among different cultivation regions. The most accurate model derived from MLP shows good prediction power for the fivefold cross validation and the test set 1 with mean square error (MSE) of 0.017 and 0.016 respectively, but a higher MSE value of 1.260 for the test set 2. After applying the AD, all models showed much lower prediction errors for the test samples within AD (IDs) than those outside the AD (ODs). MLP remains the best predictive model with an MSE value of 0.030 for the IDs. CONCLUSION: Cultivation years have a close relationship with bioactive components of AG. The constructed models and AD are also able to predict the cultivation years and discriminate samples that have inaccurate prediction results. The AD-equipped models used in this study provide useful tools for determining the age of AG in the market and are freely available at https://github.com/dreadlesss/Panax_age_predictor .

Electrostatic Field-Induced Oscillator Strength Focusing in Molecules.

The development of light-harvesting devices based on molecular materials depends critically on the ability to focus the electronic oscillator strength of molecules into the UV-vis spectral window. Typical molecular chromophores have only about 1% of their total electronic oscillator strength in this spectral region and thus perform at a small fraction of their possible effectiveness. This theoretical study finds that the electronic oscillator strength of polyenes in the UV-vis region may be enhanced by 1 order of magnitude using electrostatic fields, motivating specific experimental studies of oscillator strength focusing. We find scaling relationships between the polyene length, the intensity of the applied field, and the field-induced increase in oscillator strength that are useful for the implementation of light-harvesting strategies based on polyenes.

Where Is the Electronic Oscillator Strength? Mapping Oscillator Strength across Molecular Absorption Spectra.

The effectiveness of solar energy capture and conversion materials derives from their ability to absorb light and to transform the excitation energy into energy stored in free carriers or chemical bonds. The Thomas-Reiche-Kuhn (TRK) sum rule mandates that the integrated (electronic) oscillator strength of an absorber equals the total number of electrons in the structure. Typical molecular chromophores place only about 1% of their oscillator strength in the UV-vis window, so individual chromophores operate at about 1% of their theoretical limit. We explore the distribution of oscillator strength as a function of excitation energy to understand this circumstance. To this aim, we use familiar independent-electron model Hamiltonians as well as first-principles electronic structure methods. While model Hamiltonians capture the qualitative electronic spectra associated with π electron chromophores, these Hamiltonians mistakenly focus the oscillator strength in the fewest low-energy transitions. Advanced electronic structure methods, in contrast, spread the oscillator strength over a very wide excitation energy range, including transitions to Rydberg and continuum states, consistent with experiment. Our analysis rationalizes the low oscillator strength in the UV-vis spectral region in molecules, a step toward the goal of oscillator strength manipulation and focusing.