Enhancer–promoter interaction (EPI) is a critical component of gene regulation. Accurately predicting EPIs across diverse cell types can advance our understanding of the molecular mechanisms behind transcriptional regulation and provide valuable insights into the onset and progression of related diseases. At present, large-scale genome-wide EPI predictions typically rely on computational approaches. However, most of these methods focus on predicting EPIs within a single cell line and lack a global perspective encompassing multiple cell lines. Furthermore, they often fail to fully account for the nonlinear relationships between features, leading to suboptimal prediction accuracy. In this study, we propose KansformerEPI, a global EPI prediction model designed for multiple cell lines. The model is built on Kansformer, an encoder that integrates KAN and Transformer, effectively capturing the nonlinear relationships among various epigenetic and sequence features. We utilized KansformerEPI to achieve cross-tissue prediction of EPIs across different cell types. This approach enhances the model’s scalability, eliminating the complexity of designing separate prediction models for individual tissues. As a result, our model is applicable to various tissues, thereby reducing dependency on extensive datasets. Experimental results demonstrate that KansformerEPI surpasses existing methods such as TransEPI, TargetFinder, and SPEID in both accuracy and stability of EPI predictions across datasets including HMEC, IMR90, K562, and NHEK.

Tool-augmented Large Language Models (LLMs) have demonstrated remarkable capabilities in language understanding and generation across various domains, with notable progress in legal applications. However, existing legal LLMs still face two major challenges in legal question answering: (1) a persistent occurrence of hallucinations in generated legal advice and (2) limited effectiveness in handling ambiguous legal queries. To address both issues, this article introduces COLLT (Clarification-Oriented Legal Large language models enhanced by Tool learning), a legal LLM framework designed around a tool learning mechanism oriented to clarification. We model the clarification, tooling, and response workflow as a budgeted sequential decision. All components, including action selection, tool invocation, and response generation, are jointly optimized through multi-task instruction tuning. The COLLT framework first determines whether a user query contains ambiguity. If ambiguity is detected, the model actively guides the user to clarify their intent; if the query is clear, the model automatically invokes appropriate legal tools to improve the accuracy and reliability of its response. To support this mechanism, we propose an end-to-end training strategy enabling the model to learn how to invoke tools adaptively across various scenarios. Specifically, we construct an instruction-tuning dataset with action tags, tool tags, and optimized responses, and use it to train a series of COLLT models based on five Chinese foundation models: ChatGLM-6B, LLaMA-3-8B, InternLM-3-8B, Qwen-2.5-7B, and Baichuan-2-7B. Experimental results show that COLLT significantly outperforms baselines across nine standard legal NLP tasks. In a free-form QA evaluation based on 500 real-world legal consultation queries, COLLT achieves notable improvements in answer accuracy, legal knowledge coverage, and response appropriateness. Further visualization analysis reveals that COLLT consistently selects appropriate tools under similar intents, and statistical analysis indicates that multi-turn clarification interactions contribute to generating higher-quality responses.

Pharmacogenomics integrates tumor molecular features with drug responses to support precision oncology, but biological domain shifts and optimization conflicts hinder transfer from screened cell lines to unlabeled patient tumors. We developed TRACER (Transfer of Response Across domains via Conflict-aware End-to-end Regulation), an end-to-end framework combining supervised drug-response prediction, biologically conditioned neighbor alignment and response-protected gradient regulation. By projecting only conflicting components of the auxiliary alignment gradient, TRACER preserves the response-optimization direction learned from experimental labels. Across ModelID-disjoint, seen-drug splits, TRACER-Full achieved a global Pearson correlation of 0.5125 $\pm $ 0.0192 on per-drug z-standardized test pairs, compared with 0.5179 $\pm $ 0.0193 for Response-only, a mean difference of $-0.0054$. Mean oncogenic Hallmark neighbor consistency increased by $0.1162 \pm 0.0210$ in alignment-excluded The Cancer Genome Atlas audit samples. At the patient level, a source-validation-defined drug partition and uniform stability analysis supported four predicted pharmacological response subtypes (PRSs) in breast invasive carcinoma (BRCA), four in uterine corpus endometrial carcinoma (UCEC) and two in sarcoma (SARC). PRS-construction-excluded drugs characterized distinct predicted mechanism-level profiles, and pan-cancer indication recovery provided a lineage-aware internal consistency assessment of the patient rankings. TRACER provides an interpretable and auditable framework for response-preserving cross-domain alignment and it generates patient pharmacogenomic hypotheses for subsequent functional validation.

With the advancement of semiconductor technology, the Network-on-Chip (NoC) has become a critical architecture for communication between multiple cores. However, failures caused by factors such as manufacturing processes can degrade its performance and stability, making efficient topology reconstruction algorithms particularly important. Conventional 2D mesh reconstruction yields irregular topologies, increasing network latency and complicating system scheduling and deployment. While REmesh structures maintain topological regularity, they struggle to balance algorithmic complexity, success rates, and reconstruction costs. This paper proposes a monotonic path shift (MPS) topological reconstruction algorithm for REmesh NoCs with core-level redundancy, based on local rapid recovery. This algorithm localizes reconstruction decisions by establishing monotonic paths between failed cores and redundant cores for recovery. It incorporates region retention and local fallback mechanisms to suppress path conflicts among multiple failed cores. Theoretical analysis shows that MPS provides an upper bound on the runtime of the algorithm, significantly reducing its time complexity. Experimental results indicate that its reconstruction success rate is comparable to that of the ACTR algorithm, with both maintaining a high repair rate even under high fault density. In terms of core reuse rate, MPS achieves significant improvements over BTTR, BSTR, and ACTR, with an average increase of approximately 10% under low-fault conditions, effectively utilizing remaining computational resources. Concurrently, the algorithm substantially reduces average migration time, accelerating recovery by several orders of magnitude in large-scale low-fault scenarios and markedly lowering online recovery overhead.