In practical applications, the degradation behavior of lithium-ion batteries exhibits significant differences due to variations in operating conditions. Meanwhile, the scarcity of labeled data poses considerable challenges for capacity prediction in terms of both accuracy and generalization. To address these issues, this article proposes a cross-domain semisupervised capacity prediction framework that integrates multigranularity feature modeling with a confidence controlled pseudolabel selection mechanism. Specifically, the proposed method enhances the model’s ability to capture the granularity of nonlinear degradation trends in battery capacity, thereby improving prediction accuracy and stability. In addition, a pseudolabel learning strategy based on confidence filtering and stagewise regulation is designed to dynamically guide high-quality pseudolabels in the target domain into training, effectively reducing the risk of noisy label propagation. Experiments conducted on eight tasks across two heterogeneous battery datasets demonstrate R$^{2}$improvements of 1.3%–8.7% and Mean Absolute Error (MAE) reductions of 38%–80%, validating the practical potential of the proposed method under complex degradation scenarios.

Peng Hu, Sizhe Liu, Changjiang Gao, Xin Huang, Xue Han, Junlan Feng, Chao Deng, Shujian Huang. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

The Robot Operating System 2 (ROS) is heavily used in autonomous systems due to its large ecosystem and modular design. However, ROS remains problematic for real-time applications despite prior efforts to improve its real-time capabilities. Many of these problems are rooted in the implementation of the ROS executor, which does not support preemption or userspecified priorities. These properties are fundamental to real-time scheduling and desirable for general systems to reduce latency. ROS variants in prior work have separately supported the preemption and prioritization of callbacks but place restrictions on the application, preventing their adoption in a real-world workload. This paper addresses these deficiencies with a novel executor framework,$\operatorname{ROS}^{\text{RT}}$, which is compatible with any type of ROS application while supporting preemptive, priority-driven scheduling. Additionally, to support flexible EDF scheduling in ROS${ }^{\text {RT }}$, a custom EDF scheduler implementation is proposed using the new Linux scheduling class SCHED_EXT. ROS${ }^{\text {RT }}$s flexibility and real-time compatibility do not come at the cost of overheads: it achieves a significant decrease in publisher-tosubscriber overhead from the native ROS executor. Finally, this paper concludes with a case study performed on the Autoware Reference System, which simulates the execution of the LiDAR module in an autonomous driving application, demonstrating the ability of$\operatorname{ROS}^{\text{RT}}$at real-time scheduling in real-world scenarios.

MOTIVATION: Drug-target interaction (DTI) prediction is crucial for drug discovery, significantly reducing costs and time in experimental searches across vast drug compound spaces. While deep learning has advanced DTI prediction accuracy, challenges remain: (i) existing methods often lack generalizability, with performance dropping significantly on unseen proteins and cross-domain settings; and (ii) current molecular relational learning often overlooks subpocket-level interactions, which are vital for a detailed understanding of binding sites. RESULTS: We introduce SP-DTI, a subpocket-informed transformer model designed to address these challenges through: (i) detailed subpocket analysis using the Cavity Identification and Analysis Routine for interaction modeling at both global and local levels, and (ii) integration of pre-trained language models into graph neural networks to encode drugs and proteins, enhancing generalizability to unlabeled data. Benchmark evaluations show that SP-DTI consistently outperforms state-of-the-art models, achieving an area under the receiver operating characteristic curve of 0.873 in unseen protein settings, an 11% improvement over the best baseline. AVAILABILITY AND IMPLEMENTATION: The model scripts are available at https://github.com/Steven51516/SP-DTI.

Molecular relational learning (MRL) is crucial for understanding the interaction behaviors between molecular pairs, a critical aspect of drug discovery and development. However, the large feasible model space of MRL poses significant challenges to benchmarking, and existing MRL frameworks face limitations in flexibility and scope. To address these challenges, avoid repetitive coding efforts, and ensure fair comparison of models, we introduce FlexMol, a comprehensive toolkit designed to facilitate the construction and evaluation of diverse model architectures across various datasets and performance metrics. FlexMol offers a robust suite of preset model components, including 16 drug encoders, 13 protein sequence encoders, 9 protein structure encoders, and 7 interaction layers. With its easy-to-use API and flexibility, FlexMol supports the dynamic construction of over 70, 000 distinct combinations of model architectures. Additionally, we provide detailed benchmark results and code examples to demonstrate FlexMol’s effectiveness in simplifying and standardizing MRL model development and comparison. FlexMol is open-sourced and available at https://github.com/Steven51516/FlexMol.

Tandem mass spectrometry has played a pivotal role in advancing proteomics, enabling the analysis of protein composition in biological tissues. Many deep learning methods have been developed for \emph{de novo} peptide sequencing task, i.e., predicting the peptide sequence for the observed mass spectrum. However, two key challenges seriously hinder the further research of this important task. Firstly, since there is no consensus for the evaluation datasets, the empirical results in different research papers are often not comparable, leading to unfair comparison. Secondly, the current methods are usually limited to amino acid-level or peptide-level precision and recall metrics. In this work, we present the first unified benchmark NovoBench for \emph{de novo} peptide sequencing, which comprises diverse mass spectrum data, integrated models, and comprehensive evaluation metrics. Recent impressive methods, including DeepNovo, PointNovo, Casanovo, InstaNovo, AdaNovo and $\pi$-HelixNovo are integrated into our framework. In addition to amino acid-level and peptide-level precision and recall, we also evaluate the models' performance in terms of identifying post-tranlational modifications (PTMs), efficiency and robustness to peptide length, noise peaks and missing fragment ratio, which are important influencing factors while seldom be considered. Leveraging this benchmark, we conduct a large-scale study of current methods, report many insightful findings that open up new possibilities for future development. The benchmark is open-sourced to facilitate future research and application. The code is available at \url{https://github.com/Westlake-OmicsAI/NovoBench}.

Test data collection for a failing integrated circuit (IC) can be very expensive and time consuming. Many companies now collect a fix amount of test data regardless of the failure characteristics. As a result, limited data collection could lead to inaccurate diagnosis, while an excessive amount increases the cost not only in terms of unnecessary test data collection but also increased cost for test execution and data-storage. In this work, the objective is to develop a method for predicting the precise amount of test data necessary to produce an accurate diagnosis. By analyzing the failing outputs of an IC during its actual test, the developed method dynamically determines which failing test pattern to terminate testing, producing an amount of test data that is sufficient for an accurate diagnosis analysis. The method leverages several statistical learning techniques, and is evaluated using actual data from a population of failing chips and five standard benchmarks. Experiments demonstrate that test-data collection can be reduced by > 30% (as compared to collecting the full-failure response) while at the same time ensuring >90% diagnosis accuracy. Prematurely terminating test-data collection at fixed levels (e.g., 100 failing bits) is also shown to negatively impact diagnosis accuracy.