Metamorphic testing is an effective technique to alleviate the oracle problem in software testing, with metamorphic relations (MRs) as its core. Currently, the identification of MRs relies mainly on domain experts, which leads to low efficiency and considerable subjectivity in MR formulation. Large Language Models (LLMs) show great potential for identifying MRs automatically. In this paper, we propose a self-verificationbased prompt construction strategy for MR identification, which consists of two stages: MR identification and MR verification. Our method is applied to the thermal-hydraulic safety design program for pebble-bed high-temperature gas-cooled reactors, with deepseek-R1 as the underlying model. Experimental results demonstrate that our approach not only covers all 44 MRs identified by experts, but also discovers 24 new MRs. It significantly outperforms existing methods in terms of coverage rate, accuracy rate, and innovation rate, thereby exhibiting superior identification capability. The research results not only advance the study of MR identification, but also provide a valuable reference and inspiration for researchers.

ABSTRACT Practical problems in scientific computation that solve differential equations rarely have explicit exact solutions. Therefore, verifying the correctness of such programs has long been a challenge due to the difficulty of producing expected outputs on test cases. In this paper, the principles of metamorphic testing are applied to verify programs that solve second‐order elliptic differential equations. We present a testing process specifically tailored for the verification testing of scientific computation programs and integrate it to the process of developing scientific software. Unlike existing approaches, we formally derive metamorphic relations from the numerical models of differential equations built in development process of scientific computing programs. The experimental results clearly show that our approach is effective in detecting faults commonly found in scientific computing programs. It outperforms the fault detecting ability of the trend method, which is a traditional testing method for scientific software.

Spatial transcriptomics has become a major method for high-throughput analysis of gene expression at the current level of cells, which can directly study gene expression changes in disease cells, reveal the occurrence and development mechanism of diseases, identify potential therapeutic targets related to diseases, and provide new clues for disease diagnosis and treatment. Although there have been major breakthroughs in the analysis and acquisition of transcriptome data, there are still many challenges in how to effectively manage, share and utilize these valuable data resources in the study of human systemic diseases. To solve this problem, we are committed to building a spatial transcriptome database website related to human systemic diseases, aiming to provide reliable datasets related to multiple diseases, and provide certain data information and analysis results to accelerate the progress of human systemic diseases research. This paper will introduce the construction process of the database website in detail, and discuss its application prospects in disease research, with the aim of promoting further development and innovation in the field of human health. Here, Human-spa mainly includes 12 Human systems, 38 disease types, 55 datasets, and Human-Spa provides a very friendly web interface for visualization and dataset parsing. In conclusion, the construction of Human-Spa will provide a powerful tool and resource platform for Human disease research. Human-Spa is available for free at http://www.human-spa.cn/