Construction and validation of machine learning algorithm for predicting depression among home-quarantined individuals during the large-scale COVID-19 outbreak: based on Adaboost model.

OBJECTIVES: COVID-19 epidemics often lead to elevated levels of depression. To accurately identify and predict depression levels in home-quarantined individuals during a COVID-19 epidemic, this study constructed a depression prediction model based on multiple machine learning algorithms and validated its effectiveness. METHODS: A cross-sectional method was used to examine the depression status of individuals quarantined at home during the epidemic via the network. Characteristics included variables on sociodemographics, COVID-19 and its prevention and control measures, impact on life, work, health and economy after the city was sealed off, and PHQ-9 scale scores. The home-quarantined subjects were randomly divided into training set and validation set according to the ratio of 7:3, and the performance of different machine learning models were compared by 10-fold cross-validation, and the model algorithm with the best performance was selected from 15 models to construct and validate the depression prediction model for home-quarantined subjects. The validity of different models was compared based on accuracy, precision, receiver operating characteristic (ROC) curve, and area under the ROC curve (AUC), and the best model suitable for the data framework of this study was identified. RESULTS: The prevalence of depression among home-quarantined individuals during the epidemic was 31.66% (202/638), and the constructed Adaboost depression prediction model had an ACC of 0.7917, an accuracy of 0.7180, and an AUC of 0.7803, which was better than the other 15 models on the combination of various performance measures. In the validation sets, the AUC was greater than 0.83. CONCLUSIONS: The Adaboost machine learning algorithm developed in this study can be used to construct a depression prediction model for home-quarantined individuals that has better machine learning performance, as well as high effectiveness, robustness, and generalizability.

Impact of sarcosine on diabetic retinopathy: Findings based on weighted gene co-expression network analysis and machine learning techniques.

AIM: To quantify the association between serum sarcosine and diabetic retinopathy (DR) using weighted gene co-expression network analysis (WGCNA). METHODS: We measured serum metabolites in 69 pairs of type 2 diabetes (T2D) patients with and without DR matched by age, gender, body mass index（BMI and HbA1c, using a propensity score matching-based approach. To identify modules and metabolites linked to DR, pathway analysis was performed using WGCNA, the Kyoto Encyclopedia of Genes and Genomes and Small-Molecule Pathway Database. The association of sarcosine with DR was estimated by restricted cubic spline and conditional logistic regression models. Its joint effects with covariates on DR were also extensively examined. RESULTS: With per interquartile range elevation of sarcosine, the adjusted odds ratio (AOR) of DR significantly decreased by 67% (AOR: 0.33, 95% confidence interval [CI]: 0.19-0.58). Similar results were also found in the tertile analysis. Compared with those in the first tertile of sarcosine, the AOR significantly decreased by 54% (AOR: 0.46, 95% CI: 0.18-1.17) and 78% (AOR: 0.22, 95% CI: 0.08-0.59) for subjects in the second and third tertiles, respectively. Compared with subjects with lower sarcosine and lower HDL-C levels, those with higher sarcosine and lower HDL-C levels had the lowest odds of DR (OR: 0.13, 95% CI: 0.04, 0.43). CONCLUSIONS: Serum sarcosine was inversely related to DR, especially in T2D patients with insufficient HDL-C. This study provides insights on a possible novel target for DR precision prevention and control, as well as a better understanding of the DR mechanism.

Plasma acylcarnitine and diabetic retinopathy: A study from Eastern China.

Background and purpose: Acylcarnitines (ACars) are important for insulin resistance and type 2 diabetes (T2D). However, their roles in diabetic retinopathy (DR) remain controversial. In this study, we aimed to investigate the association of ACars with DR and their values in DR detection. Methods: This was a two-center case-control study based on the propensity score matching approach between August 2017 to June 2018 in Eastern China. Multivariable logistic regression models were applied to estimate the association of plasma ACars with DR. Differential ACars were screened by models of least absolute shrinkage and selection operator, elastic net, and weighted quantile sum regression, and their roles in DR identification were further evaluated by the area under the receiver operating curve (AUC). Results: Eight of twenty plasma ACars (8:0, 12:0, 12:1, 14:1, 16:2, 18:0, 18:2 and 18:3) were associated with DR, while only ACar 8:0 was selected by three variable selection methods. As compared to those with the 1st tertile of ACar 8:0, the adjusted odds ratio (OR) and 95% confidence interval (CI) of DR were 0.22 (0.08, 0.59) and 0.12 (0.04, 0.36) for subjects in the 2nd and 3rd tertiles, respectively (P for trend < 0.001). Consistent associations were also observed in both restricted cubic spline regression models and subgroup analyses. AUC (95% CI) were 0.74 (0.66, 0.82) for ACar 8:0 alone and 0.77 (0.70, 0.85) for ACar 8:0 combined with covariates. Conclusions: Our findings suggest higher ACar 8:0 is significantly associated with a decreased risk of DR, which provides a unique window for early identification of DR.

Phytochemicals and bioactive analysis of different sweet tea (Lithocarpus litseifolius [Hance] Chun) varieties.

This paper aims to investigate the constituent phytochemicals and biological activities of six Lithocarpus litseifolius (Hance) Chun (L. litseifolius) cultivars. Processed by UPLC-MS/MS and untargeted metabolomics analysis software (MS-DIAL and MS-FINDER), components of different L. litseifolius cultivars were analyzed qualitatively and quantitatively. By conducting a comprehensive analysis of spectra, MS/MS fragmentation, and retention time, 40 compounds (including 6 phenolic glycosides and 34 flavonoids) were identified rapidly and accurately. Moreover, the total phenolic content (TPC), total flavonoid content (TFC), antioxidant, antimicrobial effects, and pheochromocytoma (PC12) cell viability of different L. litseifolius varieties were evaluated. Purple small leaf (PSL) was found to be the most superior variety with the highest antioxidant activities (8.27, 10.09, and 10.80 mg TE/g for DPPH, ABTS, and FRAP assays, respectively), broad-spectrum antimicrobial activity, and lowest cell cytotoxicity. The results showed that these bioactivities were closely related to phytochemicals among the six cultivars. PRACTICAL APPLICATIONS: L. litseifolius is well known as an important folk drink in the Southern China folklore for its specific efficacy in preventing diabetes. However, there is still a paucity of information on L. litseifolius, as other active components and bioactivities of it remain to be investigated. The wide range of flavonoids discovered in L. litseifolius provides abundant natural health-promoting agents for further applications in medicine and functional food.

Effect of Heat Treatment on Cr2Nb Phase and Properties of Spark Plasma Sintered Cu-2Cr-1Nb Alloy.

Achieving a good match between strength and conductivity is a challenge of the development of the high-performance Cu-Cr-Nb alloy for aerospace and fusion energy. The effect of heat treatment on Cr2Nb phase, strength and conductivity of spark plasma sintered (SPSed) Cu-2Cr-1Nb (at%) alloy was investigated. The results illustrated that Cr2Nb phase of Cu-2Cr-1Nb alloy can be regulated by heat treatment, multi-scale Cr2Nb phase with sizes of 0.10-0.50 µm, 30-100 nm and less than 30 nm was obtained, and the strength and conductivity were significantly increased after heat treatment at 500 °C for 2 h, the room temperature tensile strength and conductivity were 332 MPa and 86.7% IACS, 2.5% and 34.8% higher than those of as-SPSed alloy; the tensile strength at 700 °C was 76 MPa. Increasing heat treatment temperature and time, the tensile strength of the alloy was reduced by 1.5%, 4.3% and 12.3% after heat treatment at 500 °C, 700 °C and 950 °C for 72 h. The good match between strength and conductivity of Cu-Cr-Nb alloy was obtained by reducing the content of alloying elements (Cr and Nb) and microstructure regulation. This approach can be used to prepare structural/functional materials with excellent strength and conductivity.

Transcription factor TCP20 regulates peach bud endodormancy by inhibiting DAM5/DAM6 and interacting with ABF2.

The dormancy-associated MADS-box (DAM) genes PpDAM5 and PpDAM6 have been shown to play important roles in bud endodormancy; however, their molecular regulatory mechanism in peach is unclear. In this study, by use of yeast one-hybrid screening, we isolated a TEOSINTE BRANCHED1/CYCLOIDEA/PROLIFERATING CELL FACTOR transcription factor, PpTCP20, in the peach cultivar 'Zhongyou 4' (Prunus persica var. nectarina). The protein was localized in the nucleus and was capable of forming a homodimer. Electrophoretic mobility shift assays demonstrated that PpTCP20 binds to a GCCCR element in the promoters of PpDAM5 and PpDAM6, and transient dual luciferase experiments showed that PpTCP20 inhibited the expression of PpDAM5 and PpDAM6 as the period of the release of flower bud endodormancy approached. In addition, PpTCP20 interacted with PpABF2 to form heterodimers to regulate bud endodormancy, and the content of abscisic acid decreased with the release of endodormancy. PpTCP20 also inhibited expression of PpABF2 to regulate endodormancy. Taken together, our results suggest that PpTCP20 regulates peach flower bud endodormancy by negatively regulating the expression of PpDAM5 and PpDAM6, and by interacting with PpABF2, thus revealing a novel regulatory mechanism in a perennial deciduous tree.