An improved multi-modal representation-learning model based on fusion networks for property prediction in drug discovery.

Accurate characterization of molecular representations plays an important role in the property prediction based on deep learning (DL) for drug discovery. However, most previous researches considered only one type of molecular representations, resulting in that it difficult to capture the full molecular feature information. In this study, a novel DL framework called multi-modal molecular representation learning fusion network (MMRLFN) is developed, which could simultaneously learn and integrate drug molecular features from molecular graphs and SMILES sequences. The developed MMRLFN method is composed of three complementary deep neural networks to learn various features from different molecular representations, such as molecular topology, local chemical background information, and substructures at varying scales. Eight public datasets involving various molecular properties used in drug discovery were employed to train and evaluate the developed MMRLFN. The obtained models showed better performances than the existing models based on mono-modal molecular representations. Additionally, a thorough analysis of the noise resistance and interpretability of the MMRLFN has been carried out. The generalization ability and effectiveness of the MMRLFN has been verified by case studies as well. Overall, the MMRLFN can accurately predict molecular properties and provide potentially valuable information from large datasets, thereby maximizing the possibility of successful drug discovery.

Downregulation of circ_PLXND1 inhibits tumorigenesis of non-small cell lung carcinoma via miR-1287-5p/ERBB3 axis.

BACKGROUND: Circular RNAs (circRNAs) have been reported to play vital roles in the progression of diverse human cancers, including non-small cell lung cancer (NSCLC). The purpose of this study was to explore the exact role and underlying mechanism of circ_PLXND1 in NSCLC progression. METHODS: Quantitative real-time polymerase chain reaction (qRT-PCR) assay was performed to determine the expression levels of circ_PLXND1, microRNA (miR)-1287-5p and human epidermal growth factor receptor 3 (ERBB3). The localization of circ_PLXND1 in NSCLC cells was tested by subcellular fractionation and localization assay. Cell angiogenesis, proliferation, apoptosis, migration and invasion were evaluated by tube formation assay, 5-ethynyl-2'-deoxyuridine (EdU) incorporation assay, 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay, flow cytometry and transwell assay. Dual-luciferase reporter assay was utilized to confirm the interaction between miR-1287-5p and circ_PLXND1 or ERBB3. Western blot assay was exploited to examine the expression of proteins. RESULTS: Circ_PLXND1 and ERBB3 were upregulated while miR-1287-5p was downregulated in NSCLC tissues and cells. Circ_PLXND1 was a stable circRNA and mainly located in cytoplasm. Circ_PLXND1 silencing suppressed the proliferation, angiogenesis, migration and invasion of NSCLC cells in vitro. For mechanism analysis, circ_PLXND1 could positively regulate ERBB3 expression via sponging miR-1287-5p. The inhibitory impacts of circ_PLXND1 knockdown on the malignant behaviors of NSCLC cells were overturned by miR-1287-5p inhibitor. Overexpression of miR-1287-5p repressed the malignant phenotypes of NSCLC cells by targeting ERBB3. Furthermore, circ_PLXND1 interference inhibited tumor growth in vivo. CONCLUSIONS: Circ_PLXND1 knockdown impeded NSCLC progression through modulating the miR-1287-5p/ERBB3 axis, indicating a promising molecular target for NSCLC treatment.

Effect of bifunctional acid on the porosity improvement of biomass-derived activated carbon for methylene blue adsorption.

Activated carbon (AC) with high specific surface area was prepared by using bifunctional H3PO4 agent, which led to dehydrating and activation effects through hydrothermal pretreatment and subsequent pyrolysis process. N2 adsorption and desorption isotherms of AC showed a high BET surface area of 2434 m2 g-1 and a total volume of pores (VT) of 2.0447 m3 g-1 for AC. The morphology and the chemical components of hydrochar and AC were characterized by scanning electron microscopy and Fourier transform infrared spectroscopy, which indicated that H3PO4 was benefitting for the formation of porous structure of AC. Subsequently, the effect of H3PO4 in hydrothermal pretreatment and activation process was investigated by comparative experiments. The removal and adsorption of methylene blue (MB) dye with different concentrations onto the AC were studied. The monolayer equilibrium adsorption capacity was 644 mg g-1, showing that AC has good adsorption qualities for methylene blue (MB). The adsorption balance data of MB on AC was best fitted to the Redlich-Peterson model. The adsorption kinetic data fit better to the pseudo-first-order model at low MB concentration, and the pseudo-second-order and Elovich models fit better when the MB concentration was rising.