A multi-center performance assessment for automated histopathological classification and grading of glioma using whole slide images.

Accurate pathological classification and grading of gliomas is crucial in clinical diagnosis and treatment. The application of deep learning techniques holds promise for automated histological pathology diagnosis. In this study, we collected 733 whole slide images from four medical centers, of which 456 were used for model training, 150 for internal validation, and 127 for multi-center testing. The study includes 5 types of common gliomas. A subtask-guided multi-instance learning image-to-label training pipeline was employed. The pipeline leveraged "patch prompting" for the model to converge with reasonable computational cost. Experiments showed that an overall accuracy of 0.79 in the internal validation dataset. The performance on the multi-center testing dataset showed an overall accuracy to 0.73. The findings suggest a minor yet acceptable performance decrease in multi-center data, demonstrating the model's strong generalizability and establishing a robust foundation for future clinical applications.

Parallel functional assessment of m6A sites in human endodermal differentiation with base editor screens.

N6-methyladenosine (m6A) plays important role in lineage specifications of embryonic stem cells. However, it is still difficult to systematically dissect the specific m6A sites that are essential for early lineage differentiation. Here, we develop an adenine base editor-based strategy to systematically identify functional m6A sites that control lineage decisions of human embryonic stem cells. We design 7999 sgRNAs targeting 6048 m6A sites to screen for m6A sites that act as either boosters or barriers to definitive endoderm specification of human embryonic stem cells. We identify 78 sgRNAs enriched in the non-definitive endoderm cells and 137 sgRNAs enriched in the definitive endoderm cells. We successfully validate two definitive endoderm promoting m6A sites on SOX2 and SDHAF1 as well as a definitive endoderm inhibiting m6A site on ADM. Our study provides a functional screening of m6A sites and paves the way for functional studies of m6A at individual m6A site level.

Perfluoroalkyl acids in surface sediments from the lower Yangtze River: Occurrence, distribution, sources, inventory, and risk assessment.

In this study, the occurrence, spatial distribution, potential sources, mass inventory, and ecological risk assessment of perfluoroalkyl acids (PFAAs) in surface sediments from the lower Yangtze River were investigated based on field and laboratory assays conducted in November 2019. The total concentrations of 13 target PFAAs (∑PFAAs) ranged from 13.83 to 20.33 ng/g dw, and perfluorooctanoic acid (PFOA) and perfluorooctanesulfonate (PFOS) were predominant in the surface sediments with average concentrations of 2.89 and 4.07 ng/g dw, respectively. The ∑PFAAs concentrations in pore-water ranged from 23.30 to 58.81 ng/L, and PFOA and PFOS were predominant with mean concentrations of 6.29 and 5.04 ng/L, respectively. The profiles of PFAAs composition in surface sediments showed limited difference. Results of fugacity model revealed that PFOS was in relative equilibrium, whereas PFOA exhibited a diffusion trend from sediments to water body. Correlation analysis and positive matrix factorization demonstrated that the main sources of ∑PFAAs were electroplating and fast-food packaging, degradation products and textile, mixed sources, and PFOA-based products. The mass inventory of ∑PFAAs was estimated to be 1680.72 kg, and the results of ecological risk assessments based on equilibrium partition and species sensitivity distribution methods suggested that the hazards of PFAAs in sediments to local aquatic organisms are low. However, the evaluation methods and control measures of PFAAs in surface sediments are still limited, requiring further research.