Emerging opportunities of using large language models for translation between drug molecules and indications.

A drug molecule is a substance that changes an organism's mental or physical state. Every approved drug has an indication, which refers to the therapeutic use of that drug for treating a particular medical condition. While the Large Language Model (LLM), a generative Artificial Intelligence (AI) technique, has recently demonstrated effectiveness in translating between molecules and their textual descriptions, there remains a gap in research regarding their application in facilitating the translation between drug molecules and indications (which describes the disease, condition or symptoms for which the drug is used), or vice versa. Addressing this challenge could greatly benefit the drug discovery process. The capability of generating a drug from a given indication would allow for the discovery of drugs targeting specific diseases or targets and ultimately provide patients with better treatments. In this paper, we first propose a new task, the translation between drug molecules and corresponding indications, and then test existing LLMs on this new task. Specifically, we consider nine variations of the T5 LLM and evaluate them on two public datasets obtained from ChEMBL and DrugBank. Our experiments show the early results of using LLMs for this task and provide a perspective on the state-of-the-art. We also emphasize the current limitations and discuss future work that has the potential to improve the performance on this task. The creation of molecules from indications, or vice versa, will allow for more efficient targeting of diseases and significantly reduce the cost of drug discovery, with the potential to revolutionize the field of drug discovery in the era of generative AI.

In Silico Screening of Natural Flavonoids against 3-Chymotrypsin-like Protease of SARS-CoV-2 Using Machine Learning and Molecular Modeling.

The "Long-COVID syndrome" has posed significant challenges due to a lack of validated therapeutic options. We developed a novel multi-step virtual screening strategy to reliably identify inhibitors against 3-chymotrypsin-like protease of SARS-CoV-2 from abundant flavonoids, which represents a promising source of antiviral and immune-boosting nutrients. We identified 57 interacting residues as contributors to the protein-ligand binding pocket. Their energy interaction profiles constituted the input features for Machine Learning (ML) models. The consensus of 25 classifiers trained using various ML algorithms attained 93.9% accuracy and a 6.4% false-positive-rate. The consensus of 10 regression models for binding energy prediction also achieved a low root-mean-square error of 1.18 kcal/mol. We screened out 120 flavonoid hits first and retained 50 drug-like hits after predefined ADMET filtering to ensure bioavailability and safety profiles. Furthermore, molecular dynamics simulations prioritized nine bioactive flavonoids as promising anti-SARS-CoV-2 agents exhibiting both high structural stability (root-mean-square deviation < 5 Å for 218 ns) and low MM/PBSA binding free energy (<-6 kcal/mol). Among them, KB-2 (PubChem-CID, 14630497) and 9-O-Methylglyceofuran (PubChem-CID, 44257401) displayed excellent binding affinity and desirable pharmacokinetic capabilities. These compounds have great potential to serve as oral nutraceuticals with therapeutic and prophylactic properties as care strategies for patients with long-COVID syndrome.

Comparison and summary of in silico prediction tools for CYP450-mediated drug metabolism.

The cytochrome P450 (CYP450) enzyme system is responsible for the metabolism of more than two-thirds of xenobiotics. This review summarizes reports of a series of in silico tools for CYP450 enzyme-drug interaction predictions, including the prediction of sites of metabolism (SOM) of a drug and the identification of inhibitor/substrates for CYP subtypes. We also evaluated four prediction tools to identify CYP inhibitors utilizing 52 of the most frequently prescribed drugs. ADMET Predictor and CYPlebrity demonstrated the best performance. We hope that this review provides guidance for choosing appropriate enzyme prediction tools from a variety of in silico platforms to meet individual needs. Such predictions are useful for medicinal chemists to prioritize their designed compounds for further drug discovery.

Development and test of highly accurate endpoint free energy methods. 1: Evaluation of ABCG2 charge model on solvation free energy prediction and optimization of atom radii suitable for more accurate solvation free energy prediction by the PBSA method.

Accurate estimation of solvation free energy (SFE) lays the foundation for accurate prediction of binding free energy. The Poisson-Boltzmann (PB) or generalized Born (GB) combined with surface area (SA) continuum solvation method (PBSA and GBSA) have been widely used in SFE calculations because they can achieve good balance between accuracy and efficiency. However, the accuracy of these methods can be affected by several factors such as the charge models, polar and nonpolar SFE calculation methods and the atom radii used in the calculation. In this work, the performance of the ABCG2 (AM1-BCC-GAFF2) charge model as well as other two charge models, that is, RESP (Restrained Electrostatic Potential) and AM1-BCC (Austin Model 1-bond charge corrections), on the SFE prediction of 544 small molecules in water by PBSA/GBSA was evaluated. In order to improve the performance of the PBSA prediction based on the ABCG2 charge, we further explored the influence of atom radii on the prediction accuracy and yielded a set of atom radius parameters for more accurate SFE prediction using PBSA based on the ABCG2/GAFF2 by reproducing the thermodynamic integration (TI) calculation results. The PB radius parameters of carbon, oxygen, sulfur, phosphorus, chloride, bromide and iodine, were adjusted. New atom types, on, oi, hn1, hn2, hn3, were introduced to further improve the fitting performance. Then, we tuned the parameters in the nonpolar SFE model using the experimental SFE data and the PB calculation results. By adopting the new radius parameters and new nonpolar SFE model, the root mean square error (RMSE) of the SFE calculation for the 544 molecules decreased from 2.38 to 1.05 kcal/mol. Finally, the new radius parameters were applied in the prediction of protein-ligand binding free energies using the MM-PBSA method. For the eight systems tested, we could observe higher correlation between the experiment data and calculation results and smaller prediction errors for the absolute binding free energies, demonstrating that our new radius parameters can improve the free energy calculation using the MM-PBSA method.

Phosphorylation at Ser289 Enhances the Oligomerization of Tau Repeat R2.

In tauopathies such as Alzheimer's disease (AD), aberrant phosphorylation causes the dissociation of tau proteins from microtubules. The dissociated tau then aggregates into sequent forms from soluble oligomers to paired helical filaments and insoluble neurofibrillary tangles (NFTs). NFTs is a hallmark of AD, while oligomers are found to be the most toxic form of the tau aggregates. Therefore, understanding tau oligomerization with regard to abnormal phosphorylation is important for the therapeutic development of AD. In this study, we investigated the impact of phosphorylated Ser289, one of the 40 aberrant phosphorylation sites of full-length tau proteins, on monomeric and dimeric structures of tau repeat R2 peptides. We carried out intensive replica exchange molecular dynamics simulation with a total simulation time of up to 0.1 ms. Our result showed that the phosphorylation significantly affected the structures of both the monomer and the dimer. For the monomer, the phosphorylation enhanced ordered-disordered structural transition and intramolecular interaction, leading to more compactness of the phosphorylated R2 compared to the wild-type one. As to the dimer, the phosphorylation increased intermolecular interaction and ß-sheet formation, which can accelerate the oligomerization of R2 peptides. This result suggests that the phosphorylation at Ser289 is likely to promote tau aggregation. We also observed a phosphorylated Ser289-Na+-phosphorylated Ser289 bridge in the phosphorylated R2 dimer, suggesting an important role of cation ions in tau aggregation. Our findings suggest that phosphorylation at Ser289 should be taken into account in the inhibitor screening of tau oligomerization.

Physiologically-Based Pharmacokinetics Modeling for Hydroxychloroquine as a Treatment for Malaria and Optimized Dosing Regimens for Different Populations.

Malaria is a severe parasite infectious disease with high fatality. As one of the approved treatments of this disease, hydroxychloroquine (HCQ) lacks clinical administration guidelines for patients with special health conditions and co-morbidities. This may result in improper dosing for different populations and lead them to suffer from severe side effects. One of the most important toxicities of HCQ overdose is cardiotoxicity. In this study, we built and validated a physiologically based pharmacokinetic modeling (PBPK) model for HCQ. With the full-PBPK model, we predicted the pharmacokinetic (PK) profile for malaria patients without other co-morbidities under the HCQ dosing regimen suggested by Food and Drug Administration (FDA) guidance. The PK profiles for different special populations were also predicted and compared to the normal population. Moreover, we proposed a series of adjusted dosing regimens for different populations with special health conditions and predicted the concentration-time (C-T) curve of the drug plasma concentration in these populations which include the pregnant population, elderly population, RA patients, and renal impairment populations. The recommended special population-dependent dosage regimens can maintain the similar drug levels observed in the virtual healthy population under the original dosing regimen provided by FDA. Last, we developed mathematic formulas for predicting dosage based on a patient's body measurements and two indexes of renal function (glomerular filtration rate and serum creatine level) for the pediatric and morbidly obese populations. Those formulas can facilitate personalized treatment of this disease. We hope to provide some advice to clinical practice when taking HCQ as a treatment for malaria patients with special health conditions or co-morbidities so that they will not suffer from severe side effects due to higher drug plasma concentration, especially cardiotoxicity.

A multiple-step in silico screening protocol to identify allosteric inhibitors of Spike-hACE2 binding.

While the COVID-19 pandemic continues to worsen, effective medicines that target the life cycle of SARS-CoV-2 are still under development. As more highly infective and dangerous variants of the coronavirus emerge, the protective power of vaccines will decrease or vanish. Thus, the development of drugs, which are free of drug resistance is direly needed. The aim of this study is to identify allosteric binding modulators from a large compound library to inhibit the binding between the Spike protein of the SARS-CoV-2 virus and human angiotensin-converting enzyme 2 (hACE2). The binding of the Spike protein to hACE2 is the first step of the infection of host cells by the coronavirus. We first built a compound library containing 77 448 antiviral compounds. Molecular docking was then conducted to preliminarily screen compounds which can potently bind to the Spike protein at two allosteric binding sites. Next, molecular dynamics simulations were performed to accurately calculate the binding affinity between the spike protein and an identified compound from docking screening and to investigate whether the compound can interfere with the binding between the Spike protein and hACE2. We successfully identified two possible drug binding sites on the Spike protein and discovered a series of antiviral compounds which can weaken the interaction between the Spike protein and hACE2 receptor through conformational changes of the key Spike residues at the Spike-hACE2 binding interface induced by the binding of the ligand at the allosteric binding site. We also applied our screening protocol to another compound library which consists of 3407 compounds for which the inhibitory activities of Spike/hACE2 binding were measured. Encouragingly, in vitro data supports that the identified compounds can inhibit the Spike-ACE2 binding. Thus, we developed a promising computational protocol to discover allosteric inhibitors of the binding of the Spike protein of SARS-CoV-2 to the hACE2 receptor, and several promising allosteric modulators were discovered.

In Silico Prediction of Pharmacokinetic Profile for Human Oral Drug Candidates Which Lack Clinical Pharmacokinetic Experiment Data.

BACKGROUNDS AND OBJECTIVES: In silico methods which can generate high-quality physiologically based pharmacokinetic (PBPK) models for arbitrary drug candidates are greatly needed to select developable drug candidates that escape drug attrition because of the poor pharmacokinetic profile. The purpose of this study is to develop a novel protocol to preliminarily predict the concentration profile of a target drug based on the PBPK model of a structurally similar template drug by combining two software platforms for PBPK modeling, the SimCYP simulator and ADMET Predictor. METHODS: The method was evaluated by utilizing 13 drug pairs from 18 drugs in the built-in database of the SimCYP software. All drug pairs have Tanimoto scores (TS) no less than 0.5. As each drug in a drug pair can serve as both target and template, 26 sets were studied in this work. Three versions (V1, V2 and V3) of models for the target drug were constructed by replacing the corresponding parameters of the template drug step by step with those predicted by ADMET Predictor for the target drug. V1 represents the replacement of molecular weight (MW), V2 includes the replacement of parameter MW, fraction unbound in plasma (fu), blood-to-plasma partition ratio (B/P), logarithm of the octanol-buffer partition coefficient (log Po:w) and acid dissociation constant (pKa). In V3, all above-mentioned parameters as well as human jejunum effective permeability (Peff), Vd and cytochrome P450 (CYP) metabolism parameters (Km, Vmax or CLint) are modified. Normalized root mean square error (NRMSE) was used for the evaluation of the model performance. RESULTS: We found that the performance of the three versions of the models depends on structural similarity of the drug pairs. For Group I drug pairs (TS ≤ 0.7), V2 and V3 performed better than V1 in terms of NRMSE; for Group II drug pairs (0.7 < TS ≤ 0.9), 8 out of 10 V3 models had NRMSE < 0.2, the cutoff we applied to judge whether the simulated concentration-time (C-T) curve was satisfactory or not. V3 outperformed the V1 and V2 versions. For the two drug pairs belonging to Group III (TS > 0.9), V2 outperformed V1 and V3, suggesting more unnecessary replacement can lower the performance of PBPK models. We also investigated how the prediction accuracy of ADMET Predictor as well as its collaboration with SimCYP influences the quality of PBPK models constructed using SimCYP. CONCLUSION: In conclusion, we generated practical guidance on applying two mainstream software packages, ADMET Predictor and SimCYP, to construct PBPK models for drugs or drug candidates that lack ADME parameters in model construction.