Skip to main content

ORIGINAL RESEARCH article

Front. Oncol.
Sec. Cancer Genetics
Volume 14 - 2024 | doi: 10.3389/fonc.2024.1505934

Integration of machine learning and experimental validation to identify the prognostic signature related to diverse programmed cell deaths in breast cancer

Provisionally accepted
Longpeng Li Longpeng Li 1,2Jinfeng Zhao Jinfeng Zhao 2Yaxin Wang Yaxin Wang 2Baoai Wu Baoai Wu 2Zhibin Zhang Zhibin Zhang 2Wanquan Chen Wanquan Chen 2Jirui Wang Jirui Wang 2Yue Cai Yue Cai 3*
  • 1 Department of Anesthesiology, Shanxi Province Cancer Hospital, Taiyuan, Shanxi Province, China
  • 2 Shanxi University, Taiyuan, Shanxi Province, China
  • 3 Shanxi Provincial Cancer Hospital, Taiyuan, China

The final, formatted version of the article will be published soon.

    Background: Programmed cell death (PCD) is closely related to the occurrence, development, and treatment of breast cancer. The aim of this study was to investigate the association between various programmed cell death patterns and the prognosis of breast cancer (BRCA) patients.The levels of 19 different programmed cell deaths in breast cancer were assessed by ssGSEA analysis, and these PCD scores were summed to obtain the PCDS for each sample. The relationship of PCDS with immune as well as metabolism-related pathways was explored.PCD-associated subtypes were obtained by unsupervised consensus clustering analysis, and differentially expressed genes between subtypes were analyzed. The prognostic signature (PCDRS) were constructed by the best combination of 101 machine learning algorithm combinations, and the C-index of PCDRS was compared with 30 published signatures. In addition, we analyzed PCDRS in relation to immune as well as therapeutic responses. The distribution of genes in different cells was explored by single-cell analysis and spatial transcriptome analysis. Potential drugs targeting key genes were analyzed by Cmap. Finally, the expression levels of key genes in clinical tissues were verified by RT-PCR.Results: PCDS showed higher levels in cancer compared to normal. Different PCDS groups showed significant differences in immune and metabolism-related pathways. PCDRS, consisting of seven key genes, showed robust predictive ability over other signatures in different datasets.The high PCDRS group had a poorer prognosis and was strongly associated with a cancer-promoting tumor microenvironment. The low PCDRS group exhibited higher levels of anti-cancer immunity and responded better to immune checkpoint inhibitors as well as chemotherapy-related drugs. Clofibrate and imatinib could serve as potential small-molecule complexes targeting SLC7A5 and BCL2A1, respectively. The mRNA expression levels of seven genes were upregulated in clinical cancer tissues.PCDRS can be used as a biomarker to assess the prognosis and treatment response of BRCA patients, which offers novel insights for prognostic monitoring and treatment personalization of BRCA patients.

    Keywords: breast cancer, machine learning, programmed cell death, Prognostic signature, Tumor Microenvironmen

    Received: 04 Oct 2024; Accepted: 10 Dec 2024.

    Copyright: © 2024 Li, Zhao, Wang, Wu, Zhang, Chen, Wang and Cai. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

    * Correspondence: Yue Cai, Shanxi Provincial Cancer Hospital, Taiyuan, China

    Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.