Adsorption, boiling or membrane filtration for disinfection by-product removal: How to make our drinking water safer?

Disinfection by-products (DBPs) generated in drinking water have become a global concern due to their potential harm to human health. Nevertheless, there are few studies about different point-of-use water treatments in household drinking water. The study aims to compare the effectiveness of three point-of-use water treatments: adsorption, boiling, and membrane filtration. The experimental results showed that the initial average concentration of volatile DBPs and non-volatile DBPs for tap water were 63.71 µg/L and 6.33 µg/L. The removal efficiency of DBPs for adsorption which were 75.6 % (the filter volumes from 0 L to 20 L) and 45.4 % (the filter volumes from 20 L to 50 L) during the service life of the filter element (50 L). Boiling had a high removal efficiency for volatile DBPs like trihalomethanes (THMs), haloacetaldehydes (HALs), haloacetonitriles (HANs), and haloketones (HKs) (90.5 %, 100 %, 100 %, and 100 %, respectively). However, boiling had a low removal efficiency which was 15 % in removing non-volatile DBPs like haloacetic acids (HAAs). Membrane filtration had a middle removal efficiency for THMs, HAAs, HALs, HKs, and HANs (45.3 %, 75.2 %, 46.5 %, 47.6 %, and 100 %, respectively). Through analysis of the correlation between dissolved organic matter (DOM) removal efficacy and DBP removal efficiency, it was found that the strongest correlation was observed between UV254 and DBP removal efficiency. Boiling showed a lower estimated cytotoxicity of DBPs compared to adsorption and membrane filtration. Cancer risk assessment of DBPs was below the specified risk range for three point-of-use water treatments. This study provides a reference for choosing point-of-use water treatments in household drinking water.

Prediction of circRNA-miRNA Association using Singular Value Decomposition and Graph Neural Networks.

A large number of experimental studies have shown that circRNAs can act as molecular sponges of microRNAs, interacting with miRNAs to regulate gene expression levels, thereby affecting the development of human diseases. Exploring the potential associations between circRNAs and miRNAs can help understand complex disease mechanisms. Considering that biological experiments are time-consuming and labor-intensive, this study proposes a computational model using a graph neural network and singular value decomposition (CMASG) for circRNA-miRNA association prediction. Specifically, graph neural networks are used to learn nonlinear feature representations of nodes, followed by matrix factorization algorithms to learn linear feature representations of nodes, and then combined feature representations learned from different perspectives. Finally, the lightGBM algorithm was used for circRNA-miRNA association prediction. The proposed CMASG model achieved an AUC value of 0.8804. The experimental results demonstrate the superiority and effectiveness of the CMASG model in predicting circRNA-miRNA association tasks.