Deep learning meets metabolomics: a methodological perspective.

Deep learning (DL), an emerging area of investigation in the fields of machine learning and artificial intelligence, has markedly advanced over the past years. DL techniques are being applied to assist medical professionals and researchers in improving clinical diagnosis, disease prediction and drug discovery. It is expected that DL will help to provide actionable knowledge from a variety of 'big data', including metabolomics data. In this review, we discuss the applicability of DL to metabolomics, while presenting and discussing several examples from recent research. We emphasize the use of DL in tackling bottlenecks in metabolomics data acquisition, processing, metabolite identification, as well as in metabolic phenotyping and biomarker discovery. Finally, we discuss how DL is used in genome-scale metabolic modelling and in interpretation of metabolomics data. The DL-based approaches discussed here may assist computational biologists with the integration, prediction and drawing of statistical inference about biological outcomes, based on metabolomics data.

Cytotoxic activity and molecular targets of atractylodin in cholangiocarcinoma cells.

OBJECTIVES: To evaluate the cytotoxic activity of atractylodin and its potential effects on heme oxygenase (HO)-1 production, STAT1/3 phosporylation and major NF-κB protein expression in the cholangiocarcinoma-associated cell line CL-6. METHODS: Standard MTT assay was used for accessing antiproliferative activity on CL-6 cells. Normal human embryonic fibroblast (OUMS) cell was taken as control cell line. Colony formation and wound healing assay were conducted to access the effects of atractylodin on cell proliferation and directional migration activity of CL-6 cells. Western blot was used for evaluating levels of protein expression and phosphorylation. KEY FINDINGS: Atractylodin exhibited selective cytotoxicity towards CL-6 as compared with OUMS with IC50 of 216.8 (212.4-233.8) and 351.2 (345.7-359.5) µm [median (range)], respectively. Exposure to the compound dose-dependently inhibited colony formation ability and decreased wound closure potential of CL-6 cells. Atractylodin treatment suppressed HO-1 production in CL-6 cells. It dose-dependently inhibited STAT1/3 protein phosphorylation and moderately inhibited NF-κB (p50), NF-κB (p52), and NF-κB (p65) protein expression in both dose- and time-dependent manner. CONCLUSIONS: Atractylodin exerts significant cytotoxic activity against CL-6 cells which may be linked to its suppressive effect on HO-1 production, STAT1/3 phosphorylation and expression of key NF-κB proteins.

Growth inhibitory effect of ß-eudesmol on cholangiocarcinoma cells and its potential suppressive effect on heme oxygenase-1 production, STAT1/3 activation, and NF-κB downregulation.

Cholangiocarcinoma (CCA) is a progressively fatal form of cancer originating from the malignant transformation of hepatic biliary cholangiocytes. The present study reports for the first time in vitro growth inhibitory activities of ß-eudesmol, the bioactive sesquiterpenoid present in the rhizome of Atractylodes lancea (Thunb) DC., with respect to its underlying potential effects on heme oxygenase-1 (HO-1) production, STAT1/3 phosphorylation, and NF-κB protein expression in human CCA cell line CL-6. The cytotoxic effect of ß-eudesmol on CL-6 cells was evaluated by MTT assay using normal human embryonic fibroblast (OUMS) as a control cell line. Results indicated that ß-eudesmol exhibited selective cytotoxicity towards CL-6 compared to OUMS with mean (±SD) IC50 (concentration that inhibits cell growth by 50%) values of 166.75 ± 3.69 and 240.01 ± 16.54 µmol/L, respectively. In addition, it also significantly suppressed colony forming and wound healing ability of CL-6 cells in a concentration-dependent manner. Western blot analysis indicated that ß-eudesmol treatment resulted in significant suppression of HO-1 production in CL-6 cells. Its inhibitory effects on the phosphorylation of STAT1/3 proteins and expression of NF-κB (p65 and p50) proteins were concentration-dependent. Taken together, these results suggest that ß-eudesmol exerts significant growth inhibitory activity on CL-6 cells that may be linked to its inhibitory effect on the production of HO-1, phosphorylation of STAT1/3, and expression of major NF-κB proteins.