Fangting Dong

Chlorobenzene compounds are environmentally relevant chlorinated chemicals for which efficient acute-hazard screening can complement experimental assessment. Reliable prediction of acute oral toxicity can support chemical prioritization, while conventional in vivo testing is costly and ethically constrained. Here, we compare global and structure-based stratified machine learning strategies under a common validation protocol. The fixed modeling dataset contained 540 mouse oral LD50 records, including 225 carbon-only and 315 heteroatom-containing compounds. We evaluated conventional RDKit descriptors (F1, 209 dimensions), KPGT-derived representations (F2, 2304 dimensions), and their combination (F3, 2513 dimensions) across Random Forest, LightGBM, CatBoost, and voting ensembles. A toxicity-stratified global hold-out test set comprising 20% of the data was reserved for final evaluation, while subgroup models were trained only on the corresponding subsets of the training partition. On this common test set, the global LightGBM model with F1 descriptors achieved the highest reported point estimates (accuracy = 0.889, balanced accuracy = 0.877, MCC = 0.753). Stratification did not establish a statistically significant improvement over the global strategy. SHAP analysis showed stable feature-importance rankings and subgroup-specific predictive associations, but these patterns do not establish distinct toxicity mechanisms. A Mahalanobis-distance applicability domain was used as a conservative extrapolation warning. The selected model is deployed through a public web platform for rapid screening of the study-specific higher acute-toxicity group (LD50 < 500 mg/kg) versus the at-or-above-threshold group (LD50 ≥ 500 mg/kg). The framework supports murine acute-hazard screening and prioritization; it does not constitute a complete environmental or human-health risk assessment.

Extracellular acidosis is considered as a hallmark of most human tumors, which plays an important role in promoting tumor malignant and aggressive phenotype in tumorigenesis. Acidosis and lactic acidosis can induce different responses in tumors. Previous studies have associated the response to lactic acidosis of tumors with good survival outcomes. In this study, we investigated the metabolomic changes in triple negative and luminal subtype breast cancer cell lines in response to acidosis and lactic acidosis. Our results showed that acidosis results in the reduction of cell viability and glycolysis in breast cancer cells, which is reversely correlated with the malignancy of cell lines. Under lactic acidosis, this reduction is reversed slightly. Untargeted metabolomic profiling revealed that glutaminolysis and fatty acid synthesis in cancer cells under acidosis are increased, while TCA cycle and glycolysis are decreased. Under lactic acidosis, the pentose phosphate pathway and acetate release are increased in MDA-MB-231 cells. The current results uncovered the different metabolic responses of breast cancer cells to acidosis and lactic acidosis, demonstrating the power of combined untargeted and stable isotope assisted metabolomics in comprehensive metabolomic analysis.

Hypoxia commonly occurs in cancers and is highly related with the occurrence, development and metastasis of cancer. Treatment of triple negative breast cancer remains challenge. Knowledge about the metabolic status of triple negative breast cancer cell lines in hypoxia is valuable for the understanding of molecular mechanisms of this tumor subtype to develop effective therapeutics.

The isoelectric point of shortened glucagon-like peptide-1, a promising new drug to therapy of type 2 diabetes, was determined using a new generation of capillary isoelectric focusing-whole column imaging. Urea was added as a solubilizer to enhance the stability and repeatability of shortened glucagon-like peptide-1. The isoelectric point of shortened glucagon-like peptide-1 measured by capillary isoelectric focusing-whole column imaging was 4.49, which was similar to the theoretical value of 4.40. Furthermore, the accuracy and repeatability of capillary isoelectric focusing-whole column imaging were investigated. The results showed that this approach provided good accuracy (bias less than 2%) and excellent repeatability (coefficients of variation less than 0.25%). The isoelectric focusing coupled with whole column imaging required only a few minutes for an assay. The study demonstrated that the capillary isoelectric focusing-whole column imaging is useful for the quality control and characterization of peptide and proteins because of its accuracy, repeatability, high performance, and high throughput.