Risk prediction for delayed clearance of high-dose methotrexate in pediatric hematological malignancies by machine learning.

This study aimed to establish a predictive model to identify children with hematologic malignancy at high risk for delayed clearance of high-dose methotrexate (HD-MTX) based on machine learning. A total of 205 patients were recruited. Five variables (hematocrit, risk classification, dose, SLC19A1 rs2838958, sex) and three variables (SLC19A1 rs2838958, sex, dose) were statistically significant in univariable analysis and, separately, multivariate logistic regression. The data was randomly split into a "training cohort" and a "validation cohort". A nomogram for prediction of delayed HD-MTX clearance was constructed using the three variables in the training dataset and validated in the validation dataset. Five machine learning algorithms (cart classification and regression trees, naïve Bayes, support vector machine, random forest, C5.0 decision tree) combined with different resampling methods were used for model building with five or three variables. When developed machine learning models were evaluated in the validation dataset, the C5.0 decision tree combined with the synthetic minority oversampling technique (SMOTE) using five variables had the highest area under the receiver operating characteristic curve (AUC 0.807 [95% CI 0.724-0.889]), a better performance than the nomogram (AUC 0.69 [95% CI 0.594-0.787]). The results support potential clinical application of machine learning for patient risk classification.

Machine learning to predict high-dose methotrexate-related neutropenia and fever in children with B-cell acute lymphoblastic leukemia.

Methotrexate (MTX), an antimetabolite for the treatment of leukemia, could cause neutropenia and subsequently fever, which might lead to treatment delay and affect prognosis. Here, we aimed to predict neutropenia and fever related to high-dose MTX using artificial intelligence. This study included 139 pediatric patients newly diagnosed with standard- or intermediate risk B-cell acute lymphoblastic leukemia. Fifty-seven SNPs of 16 genes were genotyped. Univariate and multivariate analysis were used to select SNPs and clinical covariates for model developing. Five machine learning algorithms combined with four resampling techniques were used to build optimal predictive model. The combination of random forest with adaptive synthetic appeared to be the best model for neutropenia (sensitivity = 0.935, specificity = 0.920, AUC = 0.927) and performed best for fever (sensitivity = 0.818, specificity = 0.924, AUC = 0.870). By machine learning, we have developed and validated comprehensive models to predict the risk of neutropenia and fever. Such models may be helpful for medical oncologists in quick decision-making.

Expression and Purification of glutathione transferase-small ubiquitin-related modifier-metallothionein fusion protein and its neuronal and hepatic protection against D-galactose-induced oxidative damage in mouse model.

The present study aimed to produce and pathophysiologically evaluate the metallothionein (MT) fusion protein. A recombinant plasmid containing DNA segment coding the pET-glutathione transferase (GST)-small ubiquitin-related modifier (SUMO)-MT fusion protein was inserted into Escherichia coli for expression. The expression level of the fusion protein was very high, reaching to 38.4% of the total supernatant proteins from the organism. Subsequent filtration through glutathione Sepharose 4B gel and Sephadex G-25 yielded an MT fusion protein with purity more than 95%. When exposed to metals, E. coli containing the GST-SUMO-MT fusion protein showed an increased accumulation of Cd(2+), Zn(2+), or Cu(2+) at approximately 4.2, 4.0, or 1.6 times higher, respectively, than those containing the control protein. Administration of GST-SUMO-MT to mice that were also treated with D-galactose to induce neuronal and hepatic damage showed a significant improvement of animal learning and memory capacity, which was depressed in mice treated by D-galactose alone. Administration of MT fusion protein also prevented D-galactose-increased malondialdehyde contents and histopathological changes in the brain and liver. Furthermore, supplement of the fusion protein significantly prevented D-galactose-increased nitric oxide contents and -decreased superoxide dismutase activity in the brain, liver, and serum. The fusion protein was also able to prevent ionizing radiation-induced DNA damage of the mouse thymus. The present study indicates that GST-SUMO-MT has a normal metal binding feature and also significantly protects the multiple tissues against oxidative damage in vivo caused by chronic exposure to D-galactose and by ionizing radiation. Therefore, GST-SUMO-MT may be a potential candidate to be developed for the clinical application.

High level expression of human endostatin in Pichia pastoris using a synthetic gene construct.

Endostatin, a 20-kDa C-terminal fragment derived from type XVIII collagen, is a potent angiogenesis inhibitor and an antitumor factor. To improve the production of recombinant human endostatin on increasing demand in clinical practice, we constructed an artificial gene encoding its mature peptide sequence in human collagen XVIII. The synthetic gene consisted of 20 codons in preference in methylotropic yeast-Pichia pastoris and was cloned into expression vector pPICZalphaA; and the recombinant protein was expressed in P. pastoris strain SMD1168 and purified to near homogeneity using heparin affinity chromatography. The amount of expressed recombinant protein in cultural media using described strategy was 80 mg/l in shake flask cultivation and 435 mg/l in high-density bioreactor fermentation. Methylthiazolium assay demonstrated that human endostatin expressed in P. pastoris using artificial synthetic gene of preference in P. pastoris was able to inhibit the acidic fibroblast growth factor-induced proliferation of endothelial cells in vitro.

High-level expression and purification of human epidermal growth factor with SUMO fusion in Escherichia coli.

Human epidermal growth factor (hEGF) can stimulate the division of various cell types and has potential clinical applications. However, the high expression of active hEGF in Escherichia coli has not been successful, as the protein contains three intra-molecular disulfide bonds that are difficult to form correctly in the bacterial intracellular environment. To solve this problem, we fused the hEGF gene with a small ubiquitin-related modifier gene (SUMO) by synthesizing an artificial SUMO-hEGF fusion gene that was highly expressed in Origami (DE3) strain. The optimal expression level of the soluble fusion protein, SUMO-hEGF, was up to 38.9% of the total cellular protein. The fusion protein was purified by Ni-NTA affinity chromatography and cleaved by a SUMO-specific protease to obtain the native hEGF, which was further purified by Ni-NTA affinity chromatography. The result of the reverse-phase HPLC showed that the purity of the recombinant cleaved hEGF was greater than 98%. The primary structure of the purified hEGF was confirmed by N-terminal amino acid sequencing and MALDI-TOF mass spectroscopy analysis. Using the method of methylthiazoletetrazolium, the mitogenic activity on Balb/c 3T3 cells of the purified hEGF was comparable to that of commercial hEGF.