Attention is all you need: utilizing attention in AI-enabled drug discovery.

Recently, attention mechanism and derived models have gained significant traction in drug development due to their outstanding performance and interpretability in handling complex data structures. This review offers an in-depth exploration of the principles underlying attention-based models and their advantages in drug discovery. We further elaborate on their applications in various aspects of drug development, from molecular screening and target binding to property prediction and molecule generation. Finally, we discuss the current challenges faced in the application of attention mechanisms and Artificial Intelligence technologies, including data quality, model interpretability and computational resource constraints, along with future directions for research. Given the accelerating pace of technological advancement, we believe that attention-based models will have an increasingly prominent role in future drug discovery. We anticipate that these models will usher in revolutionary breakthroughs in the pharmaceutical domain, significantly accelerating the pace of drug development.

Comprehensive Analysis of the Molecular Characteristics and Prognosis value of AT II-associated Genes in Non-small Cell Lung Cancer.

Alveolar type II (AT II) is a key structure of the distal lung epithelium and essential to maintain normal lung homeostasis. Dedifferentiation of AT II cells is significantly correlated with lung tumor progression. However, the potential molecular mechanism and clinical significance of AT II-associated genes for lung cancer has not yet been fully elucidated. In this study, we comprehensively analyzed the gene expression, prognosis value, genetic alteration, and immune cell infiltration of eight AT II-associated genes (AQP4, SFTPB, SFTPC, SFTPD, CLDN18, FOXA2, NKX2-1, and PGC) in Nonsmall Cell Lung Cancer (NSCLC). The results have shown that the expression of eight genes were remarkably reduced in cancer tissues and observably relating to clinical cancer stages. Survival analysis of the eight genes revealed that low-expression of CLDN18, FOXA2, NKX2-1, PGC, SFTPB, SFTPC, and SFTPD were significantly related to a reduced progression-free survival (FP), and low CLDN18, FOXA2, and SFTPD mRNA expression led to a short postprogression survival (PPS). Meanwhile, the alteration of 8 AT II-associated genes covered 273 out of 1053 NSCLC samples (26%). Additionally, the expression level of eight genes were significantly correlated with the infiltration of diverse immune cells, including six types of CD4+T cells, macrophages, neutrophils, B cells, CD8+ T cells, and dendritic cells. In summary, this study provided clues of the values of eight AT II-associated genes as clinical biomarkers and therapeutic targets in NSCLC and might provide some new inspirations to assist the design of new immunotherapies.

Development and application of ribonucleic acid therapy strategies against COVID-19.

The Coronavirus disease 2019 (COVID-19) pandemic is caused by the severe acute respiratory syndrome 2 coronavirus (SARS-CoV-2), remaining a global health crisis since its outbreak until now. Advanced biotechnology and research findings have revealed many suitable viral and host targets for a wide range of therapeutic strategies. The emerging ribonucleic acid therapy can modulate gene expression by post-transcriptional gene silencing (PTGS) based on Watson-Crick base pairing. RNA therapies, including antisense oligonucleotides (ASO), ribozymes, RNA interference (RNAi), aptamers, etc., were used to treat SARS-CoV whose genome is similar to SARV-CoV-2, and the past experience also applies for the treatment of COVID-19. Several studies against SARS-CoV-2 based on RNA therapeutic strategy have been reported, and a dozen of relevant preclinical or clinical trials are in process globally. RNA therapy has been a very active and important part of COVID-19 treatment. In this review, we focus on the progress of ribonucleic acid therapeutic strategies development and application, discuss corresponding problems and challenges, and suggest new strategies and solutions.

Identification of Helicobacter pylori Membrane Proteins Using Sequence-Based Features.

Helicobacter pylori (H. pylori) is the most common risk factor for gastric cancer worldwide. The membrane proteins of the H. pylori are involved in bacterial adherence and play a vital role in the field of drug discovery. Thus, an accurate and cost-effective computational model is needed to predict the uncharacterized membrane proteins of H. pylori. In this study, a reliable benchmark dataset consisted of 114 membrane and 219 nonmembrane proteins was constructed based on UniProt. A support vector machine- (SVM-) based model was developed for discriminating H. pylori membrane proteins from nonmembrane proteins by using sequence information. Cross-validation showed that our method achieved good performance with an accuracy of 91.29%. It is anticipated that the proposed model will be useful for the annotation of H. pylori membrane proteins and the development of new anti-H. pylori agents.

Analysis and modeling of myopia-related factors based on questionnaire survey.

With the rapid development of science and technology, the trend of low age myopia is becoming increasingly significant. The latest national survey done by the Chinese government found that more than 80% of Chinese teenagers suffer from myopia. Adolescent myopia is closely related to living environment, heredity, and living habits. Quantifying the relationship between myopia and living environment, heredity, and living habits is conductive to the prevention and intervention of adolescent myopia. In this study, we investigated the relationships between four main factors (environment, habits, parental vision, and demographic) and myopia status by analyzing the questionnaire data. Data were collected from Chengdu, China in 2021, including 2808 myopia samples and 5693 non-myopia samples, with a total of 22 features. Then, these 22 features were inputted into three machine learning algorithms to discriminate the two classes of samples. Results show that the computational model could produce an AUC of 0.768. To pick out the most important features which play important roles in classification, we used incremental feature selection strategy to screen the 22 features. As a result, we found that the 4 most influential features with XGBoost could achieve a competitive AUC of 0.764. To further investigate the risk and protective factors affecting adolescent myopia, we used OR values derived from MLE-LR to analyze the relationship between 22 features and adolescent myopia. Results showed that the age variable was the most significant risk factor for myopia, followed by the myopia status of parents. The most protective factor for eyesight is the measure taken by the children, followed by the distance between books and eyes when reading. These discoveries can guide the prevention and control of myopia in children and adolescents.