Graph neural networks (GNN) has been considered as an attractive modelling method for molecular property prediction, and numerous studies have shown that GNN could yield more promising results than traditional descriptor-based methods. In this study, based on 11 public datasets covering various property endpoints, the predictive capacity and computational efficiency of the prediction models developed by eight machine learning (ML) algorithms, including four descriptor-based models (SVM, XGBoost, RF and DNN) and four graph-based models (GCN, GAT, MPNN and Attentive FP), were extensively tested and compared. The results demonstrate that on average the descriptor-based models outperform the graph-based models in terms of prediction accuracy and computational efficiency. SVM generally achieves the best predictions for the regression tasks. Both RF and XGBoost can achieve reliable predictions for the classification tasks, and some of the graph-based models, such as Attentive FP and GCN, can yield outstanding performance for a fraction of larger or multi-task datasets. In terms of computational cost, XGBoost and RF are the two most efficient algorithms and only need a few seconds to train a model even for a large dataset. The model interpretations by the SHAP method can effectively explore the established domain knowledge for the descriptor-based models. Finally, we explored use of these models for virtual screening (VS) towards HIV and demonstrated that different ML algorithms offer diverse VS profiles. All in all, we believe that the off-the-shelf descriptor-based models still can be directly employed to accurately predict various chemical endpoints with excellent computability and interpretability.

Amino acids are the building blocks of protein synthesis. They are structural elements and energy sources of cells necessary for normal cell growth, differentiation and function. Amino acid metabolism disorders have been linked with a number of pathological conditions, including metabolic diseases, cardiovascular diseases, immune diseases, and cancer. In the case of tumors, alterations in amino acid metabolism can be used not only as clinical indicators of cancer progression but also as therapeutic strategies. Since the growth and development of tumors depend on the intake of foreign amino acids, more and more studies have targeted the metabolism of tumor-related amino acids to selectively kill tumor cells. Furthermore, immune-related studies have confirmed that amino acid metabolism regulates the function of effector T cells and regulatory T cells, affecting the function of immune cells. Therefore, studying amino acid metabolism associated with disease and identifying targets in amino acid metabolic pathways may be helpful for disease treatment. This article mainly focuses on the research of amino acid metabolism in tumor-oriented diseases, and reviews the research and clinical research progress of metabolic diseases, cardiovascular diseases and immune-related diseases related to amino acid metabolism, in order to provide theoretical basis for targeted therapy of amino acid metabolism.

N6-methyladenosine (m6A) methylation, a well-known modification with new epigenetic functions, has been reported to participate in the tumorigenesis of hepatocellular carcinoma (HCC), providing novel insights into the molecular pathogenesis of this disease. However, as the key component of m6A methylation, Wilms tumor 1-associated protein (WTAP) has not been well studied in HCC. Here we investigated the biological role and underlying mechanism of WTAP in liver cancer. We determined the expression of WTAP and its correlation with clinicopathological features using tissue microarrays and the Cancer Genome Atlas (TCGA) dataset. And we clarified the effects of WTAP on HCC cells using cell proliferation assay, colony formation, Edu assay and subcutaneous xenograft experiments. We then applied RNA sequencing combined with gene expression omnibus (GEO) data to screen candidate targets of WTAP. Finally, we investigated the regulatory mechanism of WTAP in HCC by m6A dot blot assay, methylated RNA immunoprecipitation (MeRIP) assay, dual luciferase reporter assay, RNA immunoprecipitation (RIP) assay and Chromatin immunoprecipitation (ChIP) assay. We demonstrated that WTAP was highly expressed in HCC which indicated the poor prognosis, and that WTAP expression served as an independent predictor of HCC survival. Functionally, WTAP promoted the proliferation capability and tumor growth of HCC cells in vitro and in vivo. Furthermore, ETS proto-oncogene 1 (ETS1) was identified as the downstream effector of WTAP. The m6A modification regulated by WTAP led to post-transcriptional suppression of ETS1, with the implication of Hu-Antigen R (HuR) as an RNA stabilizer. Then ETS1 was found to inhibit the progression of HCC and could rescue the phenotype induced by WTAP deficiency. Moreover, WTAP modulated the G2/M phase of HCC cells through a p21/p27-dependent pattern mediated by ETS1. We have identified that WTAP is significantly up-regulated in HCC and promotes liver cancer development. WTAP-guided m6A modification contributes to the progression of HCC via the HuR-ETS1-p21/p27 axis. Our study is the first to report that WTAP-mediated m6A methylation has a crucial role in HCC oncogenesis, and highlights WTAP as a potential therapeutic target of HCC treatment.

Background: TikTok was an important channel for consumers to access and adopt health information. But the quality of health content in TikTok remains underinvestigated. Objective: Our study aimed to identify upload sources, contents, and feature information of gallstone disease videos on TikTok and further evaluated the factors related to video quality. Methods: We investigated the first 100 gallstone-related videos on TikTok and analyzed these videos’ upload sources, content, and characteristics. The quality of videos was evaluated using quantitative scoring tools such as DISCERN instrument, the Journal of American Medical Association (JAMA) benchmark criteria, and Global Quality Scores (GQS). Moreover, the correlation between video quality and video characteristics, including duration, likes, comments, and shares, was further investigated. Results: According to video sources, 81% of the videos were posted by doctors. Furthermore, disease knowledge was the most dominant video content, accounting for 56% of all the videos. The mean DISCERN, JAMA, and GQS scores of all 100 videos are 39.61 (SD 11.36), 2.00 (SD 0.40), and 2.76 (SD 0.95), respectively. According to DISCERN and GQS, gallstone-related videos’ quality score on TikTok is not high, mainly at fair (43/100, 43%,) and moderate (46/100, 46%). The total DISCERN scores of doctors were significantly higher than that of individuals and news agencies, surgery techniques were significantly higher than lifestyle and news, and disease knowledge was significantly higher than news, respectively. DISCERN scores and video duration were positively correlated. Negative correlations were found between DISCERN scores and likes and shares of videos. In GQS analysis, no significant differences were found between groups based on different sources or different contents. JAMA was excluded in the video quality and correlation analysis due to a lack of discrimination and inability to evaluate the video quality accurately. Conclusions: Although the videos of gallstones on TikTok are mainly provided by doctors and contain disease knowledge, they are of low quality. We found a positive correlation between video duration and video quality. High-quality videos received low attention, and popular videos were of low quality. Medical information on TikTok is currently not rigorous enough to guide patients to make accurate judgments. TikTok was not an appropriate source of knowledge to educate patients due to the low quality and reliability of the information.

Background: N6-methyladenosine (m⁶A) modification is an emerging layer of epigenetic regulation which is widely implicated in the tumorigenicity of hepatocellular carcinoma (HCC), offering a novel perspective for investigating molecular pathogenesis of this disease. The role of AlkB homolog 5 (ALKBH5), one of the m⁶A demethylases, has not been fully explored in HCC. Here we clarify the biological profile and potential mechanisms of ALKBH5 in HCC. Methods: Expression of ALKBH5 and its correlation with clinicopathological characteristics of HCC were evaluated using tissue microarrays and online datasets. And biological effects of ALKBH5 in HCC were determined in vitro and in vivo. Subsequently, methylated RNA immunoprecipitation sequencing (MeRIP-seq) combined with RNA sequencing (RNA-seq), and following m⁶A dot blot, MeRIP-qPCR, RIP-qPCR or dual luciferase reporter assays were employed to screen and validate the candidate targets of ALKBH5. Results: We demonstrated that ALKBH5 was down-regulated in HCC, and decreased ALKBH5 expression was an independent prognostic factor of worse survival in HCC patients. Functionally, ALKBH5 suppressed the proliferation and invasion capabilities of HCC cells in vitro and in vivo. Mechanistically, ALKBH5-mediated m⁶A demethylation led to a post-transcriptional inhibition of LY6/PLAUR Domain Containing 1 (LYPD1), which could be recognized and stabilized by the m⁶A effector IGF2BP1. In addition, we identified that LYPD1 induced oncogenic behaviors of tumors in contrast to ALKBH5. Dysregulation of ALKBH5/LYPD1 axis impelled the progression of HCC. Conclusion: Our study reveals that ALKBH5, characterized as a tumor suppressor, attenuates the expression of LYPD1 via an m⁶A-dependent manner in HCC cells. Our findings enrich the landscape of m⁶A-modulated tumor malignancy, and provide new insights into potential biomarkers and therapeutic targets of HCC treatment.

Proteolysis-targeting chimeras (PROTACs), which selectively degrade targeted proteins by the ubiquitin-proteasome system, have emerged as a novel therapeutic technology with potential advantages over traditional inhibition strategies. In the past few years, this technology has achieved substantial progress and two PROTACs have been advanced into phase I clinical trials. However, this technology is still maturing and the design of PROTACs remains a great challenge. In order to promote the rational design of PROTACs, we present PROTAC-DB, a web-based open-access database that integrates structural information and experimental data of PROTACs. Currently, PROTAC-DB consists of 1662 PROTACs, 202 warheads (small molecules that target the proteins of interest), 65 E3 ligands (small molecules capable of recruiting E3 ligases) and 806 linkers, as well as their chemical structures, biological activities, and physicochemical properties. Except the biological activities of warheads and E3 ligands, PROTAC-DB also provides the degradation capacities, binding affinities and cellular activities for PROTACs. PROTAC-DB can be queried with two general searching approaches: text-based (target name, compound name or ID) and structure-based. In addition, for the convenience of users, a filtering tool for the searching results based on the physicochemical properties of compounds is also offered. PROTAC-DB is freely accessible at http://cadd.zju.edu.cn/protacdb/.