A group of theobromine derivatives was designed based on the key pharmacophoric characteristics of VEGFR-2 inhibitors. HepG2 and MCF-7 cancer cell lines were used to test the obtained compounds for their in vitro anti-proliferative activities. Compound 15 (2-(3,7-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-1-yl)-N-(4-(1-(2-(4-hydroxybenzoyl)hydrazono)ethyl) phenyl)acetamide) was the most potent cytotoxic member against MCF-7 (IC50 = 0.42 µM) and HepG2 (IC50 = 0.22 µM). The effectiveness of VEGFR-2 inhibition was assessed for compound 15, and its IC50 value was calculated to be 0.067 µM. Additional cellular mechanistic investigations showed that compound 15 dramatically increased the population of apoptotic HepG2 cells in both early and late apoptosis. The investigation of apoptotic markers confirmed that compound 15 upregulated the levels of BAX (2.26-fold) and downregulated the levels of Bcl-2 (4.4-fold). The molecular docking investigations, MM-GPSA, PLIP studies, and MD simulations validated the potential of compound 15 to be a VEGFR-2 inhibitor. DFT calculations have been completed to comprehend how the electrical charge is distributed within compound 15 and to predict how it would bond to VEGFR-2. Lastly, ADMET prediction showed that the designed members have drug-like characteristics and minimal levels of toxicity. In conclusion, our in vitro and in silico investigations showed that compound 15 exhibited promising apoptotic anticancer potential through the suppression of VEGFR-2.

Predicting protein stability change upon variation through a computational approach is a valuable tool to unveil the mechanisms of mutation-induced drug failure and develop immunotherapy strategies. Some previous machine learning-based techniques exhibit anti-symmetric bias toward destabilizing situations, whereas others struggle with generalization to unseen examples. To address these issues, we propose a gated graph neural network-based approach to predict changes in protein stability upon mutation. The model uses message passing to encode the links between the molecular structure and property after eliminating the non-mutant structure and creating input feature vectors. While doing so, it also incorporates the coordinates of the raw atoms to provide spatial insights into the chemical systems. We test the model on the Ssym, Myoglobin, Broom, and p53 datasets to demonstrate the generalization performance. Compared to existing approaches, our proposed method achieves improved linearity with symmetry in less time. The code for this study is available at: https://github.com/HongzhouTang/Pros-GNN.

Tuberculosis (TB), caused by Mycobacterium tuberculosis (MTB), remains a major threat to global health, with the emergence of multi-drug and extensively drug-resistant strains posing a serious challenge. Thereby, understanding the molecular basis of MTB virulence and disease pathogenesis is critical for developing effective therapeutic strategies. Targeting proteins involved in central metabolism has been recognized as a promising therapeutic approach to combat MTB. In this regard, the enzyme AckA of the acetate metabolic pathway which produces acetate from acetyl phosphate, is an important drug target for various pathogenic organisms. Therefore, this study aimed to identify potential AckA inhibitors through in silico methods, including molecular modeling, molecular dynamics simulation (MDS), and high-throughput virtual screening (HTVS) followed by ADMETox, MMGBSA, Density Functional Theory (DFT) calculations. HTVS of one million compounds from the ZINC database against AckA resulted in the top five hits (ZINC82048449, ZINC1219737510, ZINC1771921358, ZINC119699567, and ZINC1427100376) with better binding affinity and optimal binding free energy. MDS studies on complexes revealed that key residues, Asn195, Asp266, Phe267, Gly314, and Asn318 played a significant role in stable interactions of the top-ranked compounds to AckA. These outcomes provide insights into the optimal binding of the leads to inhibit the acetate pathway and aid in the rational design of novel therapeutic agents. Thus, the identified leads may act as promising compounds for targeting AckA and may serve as a potential therapeutic modality for treating TB. Our findings offer valuable insights into the inhibition of the acetate pathway, while also serving as a blueprint for rational drug design. The identified leads hold promise as compelling compounds for targeting AckA, thereby offering a potential therapeutic avenue for tackling TB. Thus, our study uncovers a pathway toward promising TB therapeutics by elucidating AckA inhibitors. By leveraging in silico methodologies, potent compounds that hold the potential to thwart AckA's role in MTB's acetate pathway have been unveiled. This breakthrough fosters optimism in the quest for novel and effective TB treatments, addressing a global health challenge with renewed vigor.

The coronavirus (COVID-19) has mutated into several variants, and evidence says that new variants are more transmissible than existing variants. Even with full-scale vaccination efforts, the theoretical threshold for eradicating COVID-19 appears out of reach. This article proposes an artificial intelligence(AI) based intelligent prediction model called Deep-SIQRV(Susceptible-Infected-Quarantined-Recovered-Vaccinated) to simulate the spreading of COVID-19. While many models assume that vaccination provides lifetime protection, we focus on the impact of waning immunity caused by the conversion of vaccinated individuals back to susceptible ones. Unlike existing models, which assume that all coronavirus-infected individuals have the same infection rate, the proposed model considers the various infection rates to analyze transmission laws and trends. Next, we consider the influence of prevention and control strategies, such as media marketing and law enforcement, on the spread of the epidemic. We employed the PAN-LDA model to extract features from COVID-19-related discussions on social media and online news articles. Moreover, the Long Short Term Memory(LSTM) model and Evolution Strategies(ES) are used to optimize transmission rates of infection and other model parameters, respectively. The experimental results on epidemic data from various Indian states demonstrate that persons infected with coronavirus had a more significant infection rate within four to nine days after infection, which corresponds to the actual transmission laws of the epidemic. The experimental results show that the proposed model has good prediction ability and obtains the Mean Absolute Percentage Error(MAPE) of 0.875%, 0.965%, 0.298%, and 0.215% for the next eight days in Maharashtra, Kerala, Karnataka, and Delhi, respectively. Our findings highlight the significance of using vaccination data, COVID-19-related posts, and information generated by the government's tremendous efforts in the prediction calculation process.