No abstract

A liquid biopsy is currently an interesting tool for measuring tumor material with the advantage of being non-invasive. The overexpression of vimentin and ezrin genes was associated with epithelial-mesenchymal transition (EMT), a key process in metastasis and progression in osteosarcoma (OS). In this study, we identified other OS-specific genes by calculating differential gene expression using the Gene Expression Omnibus (GEO) database, confirmed by using quantitative reverse transcription-PCR (qRT-PCR) to detect OS-specific genes, including VIM and ezrin in the buffy coat, which were obtained from the whole blood of OS patients and healthy donors. Furthermore, the diagnostic model for OS detection was generated by utilizing binary logistic regression with a multivariable fractional polynomial (MFP) algorithm. The model incorporating VIM, ezrin, and COL5A2 genes exhibited outstanding discriminative ability, as determined by the receiver operating characteristic curve (AUC = 0.9805, 95% CI 0.9603, 1.000). At the probability cut-off value of 0.3366, the sensitivity and the specificity of the model for detecting OS were 98.63% (95% CI 90.5, 99.7) and 94.94% (95% CI 87.5, 98.6), respectively. Bioinformatic analysis and qRT-PCR, in our study, identified three candidate genes that are potential diagnostic and prognostic genes for OS.

This study applies statistical methods and deep learning techniques to forecast the daily average PM 2.5 concentration in northern Thailand, where the concentration is usually high and exceeds the safe level. The data used in the analysis are collected from January 2018 to December 2020 from 16 air monitoring stations. The statistical methods used are Holt-Winters exponential smoothing (ETS), autoregressive integrated moving average (ARIMA), and dynamic linear model (DLM). The deep learning techniques considered in this study are the recurrent neural network (RNN) and long-short term memory (LSTM). To compare the predictive performance of both methods, we use the root mean square error (RMSE). The result indicates that statistical methods, especially ARIMA, perform better than the deep learning techniques in most stations. Moreover, LSTM tends to provide higher accuracy than the RNN, especially with more number of nodes.

No abstract

No abstract