Utilizing Graph Neural Networks for Breast Cancer Prognosis Prediction with High-dimensional Genomic Data.

An accurate prediction of breast cancer is essential to help physicians make appropriate treatment recommendations to reduce the chance of excessive treatment, avoiding unnecessary anxiety for patients. Cancer prognosis is highly related to patients' genomic features, which are high-dimensional in nature. In this study, we utilize a systems biology feature selector for dimension reduction to select 20 prognostic biomarkers that are considered closely related to breast cancer prognosis from the high dimensional RNA Sequencing (RNA-Seq) data. Furthermore, we establish a graph neural network (GNN) and a multi-layer perception (MLP) graph-level readout method to better extract the underlying gene interactions from the corresponding gene interaction network (GIN). With the help of GINs, the model performs the best among all baseline models, especially in the area under the precision-recall curve (AUPRC) by as large as 23%. The results demonstrate that our approach using GNNs can successfully extract high-dimensional and complicated interactions within genomic data.

Sonoelectrochemical nitrided graphene nanosheets with vacancies and their applications for catalysis and sensing of uric acid oxidation.

A sonoelectrochemical method for preparing N-doped defective graphene nanosheets (N/O-dGNs) with point defects and 5-9 or 5-8-5 vacancies and oxygen-containing groups was successfully demonstrated. In this one-pot approach, the N-bonding configuration and N content of N/O-dGNs were finely tuned by the ultrasonic power (192, 320, and 640 W). The N content in atomic percentage (at%) for N/O-dGN (N/O-dGN320W) with point defects and 5-8-5 vacancy prepared at 320 W power was 5.6 at%, greater than 3.0 at% and 2.6 at% for N/O-dGN with point defects and 5-9 vacancies at 192 W and 640 W power (N/O-dGN192W and N/O-dGN640W), respectively. N-bonding sites on N/O-dGN320W were dominantly amine N (2.1 at%) and pyrrolic N (2.4 at%). Additionally, the electrocatalytic activity of N/O-dGN192W, N/O-dGN320W, and N/O-dGN640W was successfully demonstrated for the sequential uric acid (UA) oxidation reaction (UOR), in which N/O-dGN320W displayed a significant mass activity (2.51 A/g). As in the transient catalysis of UOR, N/O-dGN320W with amine N showed 400.8 µA mM-1 cm-2 in sensitivity within a wide linear analysis range (1.5 × 10-2-6 mM) for amperometrically sensing UA. The results of real sample experiments using serum samples further demonstrated the potential of N/O-dGN320W as a non-enzymatic UA sensor.

Numerical Evidence for a Haagerup Conformal Field Theory.

We numerically study an anyon chain based on the Haagerup fusion category and find evidence that it leads in the long-distance limit to a conformal field theory whose central charge is ∼2. Fusion categories generalize the concept of finite group symmetries to noninvertible symmetry operations, and the Haagerup fusion category is the simplest one which comes from neither finite groups nor affine Lie algebras. As such, ours is the first example of conformal field theories which have truly exotic generalized symmetries. Basically the same result was independently obtained in the preceding Letter [Phys. Rev. Lett. 128, 231602 (2022)PRLTAO0031-900710.1103/PhysRevLett.128.231602].