The final, formatted version of the article will be published soon.
ORIGINAL RESEARCH article
Front. Immunol.
Sec. Cancer Immunity and Immunotherapy
Volume 15 - 2024 |
doi: 10.3389/fimmu.2024.1497300
This article is part of the Research Topic Community Series in Novel Reliable Approaches for Prediction and Clinical Decision-making in Cancer: Volume II View all articles
Multi-omics characterization and machine learning of lung adenocarcinoma molecular subtypes to guide precise chemotherapy and immunotherapy
Provisionally accepted- 1 Guang'an People's Hospital, Guang'an, China
- 2 People’s Hospital of Deyang City, Deyang, Sichuan Province, China
- 3 The First People’s Hospital of Lianyungang, Lianyungang, Jiangsu Province, China
Lung adenocarcinoma (LUAD) is a heterogeneous tumor characterized by diverse genetic and molecular alterations. Developing a multi-omics-based classification system for LUAD is urgently needed to advance biological understanding. Data on clinical and pathological characteristics, genetic alterations, DNA methylation patterns, and the expression of mRNA, lncRNA, and microRNA, along with somatic mutations in LUAD patients, were gathered from the TCGA and GEO datasets. A computational workflow was utilized to merge multi-omics data from LUAD patients through 10 clustering techniques, which were paired with 10 machine learning methods to pinpoint detailed molecular subgroups and refine a prognostic risk model. The disparities in somatic mutations, copy number alterations, and immune cell infiltration between highrisk and low-risk groups were assessed. The effectiveness of immunotherapy in patients was evaluated through the TIDE and SubMap algorithms, supplemented by data from various immunotherapy groups. Furthermore, the Cancer Therapeutics Response Portal (CTRP) and the PRISM Repurposing dataset (PRISM) were employed to investigate new drug treatment approaches for LUAD. In the end, the role of SLC2A1 in tumor dynamics was examined using RT-PCR, immunohistochemistry, CCK-8, wound healing, and transwell tests. By employing multi-omics clustering, we discovered two unique cancer subtypes linked to prognosis, with CS2 demonstrating a better outcome.A strong model made up of 17 genes was created using a random survival forest (RSF) method, which turned out to be an independent predictor of overall survival and showed reliable and impressive performance. The low-risk group not only had a better prognosis but was also more likely to display the "cold tumor" phenotype. On the other hand, individuals in the high-risk group showed a worse outlook and were more likely to respond positively to immunotherapy and six particular chemotherapy medications.Laboratory cell tests demonstrated that SLC2A1 is abundantly present in LUAD tissues and cells, greatly enhancing the proliferation and movement of LUAD cells. Thorough examination of multi-omics data offers vital understanding and improves the molecular categorization of LUAD. Utilizing a powerful machine learning system, we highlight the immense potential of the riskscore in providing individualized risk evaluations and customized treatment suggestions for LUAD patients.
Keywords: Lung Adenocarcinoma, machine learning, Multi-omics analysis, prognosis, Treatment
Received: 16 Sep 2024; Accepted: 08 Nov 2024.
Copyright: © 2024 Zhang, Wang and Qian. 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:
Yi Zhang, Guang'an People's Hospital, Guang'an, China
Haitao Qian, The First People’s Hospital of Lianyungang, Lianyungang, 222002, Jiangsu Province, 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.