1
|
Qian X, Yang G, Li F, Zhang X, Zhu X, Lai X, Xiao X, Wang T, Wang J. DeepLION2: deep multi-instance contrastive learning framework enhancing the prediction of cancer-associated T cell receptors by attention strategy on motifs. Front Immunol 2024; 15:1345586. [PMID: 38515756 PMCID: PMC10956474 DOI: 10.3389/fimmu.2024.1345586] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/28/2023] [Accepted: 02/19/2024] [Indexed: 03/23/2024] Open
Abstract
Introduction T cell receptor (TCR) repertoires provide valuable insights into complex human diseases, including cancers. Recent advancements in immune sequencing technology have significantly improved our understanding of TCR repertoire. Some computational methods have been devised to identify cancer-associated TCRs and enable cancer detection using TCR sequencing data. However, the existing methods are often limited by their inadequate consideration of the correlations among TCRs within a repertoire, hindering the identification of crucial TCRs. Additionally, the sparsity of cancer-associated TCR distribution presents a challenge in accurate prediction. Methods To address these issues, we presented DeepLION2, an innovative deep multi-instance contrastive learning framework specifically designed to enhance cancer-associated TCR prediction. DeepLION2 leveraged content-based sparse self-attention, focusing on the top k related TCRs for each TCR, to effectively model inter-TCR correlations. Furthermore, it adopted a contrastive learning strategy for bootstrapping parameter updates of the attention matrix, preventing the model from fixating on non-cancer-associated TCRs. Results Extensive experimentation on diverse patient cohorts, encompassing over ten cancer types, demonstrated that DeepLION2 significantly outperformed current state-of-the-art methods in terms of accuracy, sensitivity, specificity, Matthews correlation coefficient, and area under the curve (AUC). Notably, DeepLION2 achieved impressive AUC values of 0.933, 0.880, and 0.763 on thyroid, lung, and gastrointestinal cancer cohorts, respectively. Furthermore, it effectively identified cancer-associated TCRs along with their key motifs, highlighting the amino acids that play a crucial role in TCR-peptide binding. Conclusion These compelling results underscore DeepLION2's potential for enhancing cancer detection and facilitating personalized cancer immunotherapy. DeepLION2 is publicly available on GitHub, at https://github.com/Bioinformatics7181/DeepLION2, for academic use only.
Collapse
Affiliation(s)
- Xinyang Qian
- School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an, China
- Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Guang Yang
- Department of Clinical Oncology, The Second Affiliated Hospital of Air Force Medical University, Xi’an, China
| | - Fan Li
- School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an, China
- Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Xuanping Zhang
- School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an, China
- Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Xiaoyan Zhu
- School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an, China
- Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Xin Lai
- School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an, China
- Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Xiao Xiao
- Genomics Institute, Geneplus-Shenzhen, Shenzhen, China
| | - Tao Wang
- Department of Thoracic Surgery, The Second Affiliated Hospital of Air Force Medical University, Xi’an, China
| | - Jiayin Wang
- School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an, China
- Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| |
Collapse
|
2
|
Joffe D, Bhatti S, Banner L, Zaya R, Gleason L, Mishra A, Kirsch I, Porcu P, Nikbakht N. The overlap of skin and blood T-cell clones in early-stage mycosis fungoides. Blood Adv 2023; 7:6206-6210. [PMID: 37611154 PMCID: PMC10582838 DOI: 10.1182/bloodadvances.2023010867] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/31/2023] [Revised: 08/16/2023] [Accepted: 08/16/2023] [Indexed: 08/25/2023] Open
Affiliation(s)
- Daniel Joffe
- Department of Dermatology and Cutaneous Biology, Thomas Jefferson University, Philadelphia, PA
| | - Safiyyah Bhatti
- Department of Dermatology and Cutaneous Biology, Thomas Jefferson University, Philadelphia, PA
- Department of Medical Oncology, Thomas Jefferson University, Philadelphia, PA
| | - Lauren Banner
- Department of Dermatology and Cutaneous Biology, Thomas Jefferson University, Philadelphia, PA
| | - Romsin Zaya
- Department of Dermatology and Cutaneous Biology, Thomas Jefferson University, Philadelphia, PA
| | - Laura Gleason
- Department of Dermatology and Cutaneous Biology, Thomas Jefferson University, Philadelphia, PA
| | - Anjali Mishra
- Department of Medical Oncology, Thomas Jefferson University, Philadelphia, PA
| | | | - Pierluigi Porcu
- Department of Medical Oncology, Thomas Jefferson University, Philadelphia, PA
| | - Neda Nikbakht
- Department of Dermatology and Cutaneous Biology, Thomas Jefferson University, Philadelphia, PA
| |
Collapse
|
3
|
Xu Y, Qian X, Zhang X, Lai X, Liu Y, Wang J. DeepLION: Deep Multi-Instance Learning Improves the Prediction of Cancer-Associated T Cell Receptors for Accurate Cancer Detection. Front Genet 2022; 13:860510. [PMID: 35601486 PMCID: PMC9121378 DOI: 10.3389/fgene.2022.860510] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/23/2022] [Accepted: 02/23/2022] [Indexed: 01/21/2023] Open
Abstract
Recent studies highlight the potential of T cell receptor (TCR) repertoires in accurately detecting cancers via noninvasive sampling. Unfortunately, due to the complicated associations among cancer antigens and the possible induced T cell responses, currently, the practical strategy for identifying cancer-associated TCRs is the computational prediction based on TCR repertoire data. Several state-of-the-art methods were proposed in recent year or two; however, the prediction algorithms were still weakened by two major issues. To facilitate the computational processes, the algorithms prefer to decompose the original TCR sequences into length-fixed amino acid fragments, while the first dilemma comes as the lengths of cancer-associated motifs are suggested to be various. Moreover, the correlations among TCRs in the same repertoire should be further considered, which are often ignored by the existing methods. We here developed a deep multi-instance learning method, named DeepLION, to improve the prediction of cancer-associated TCRs by considering these issues. First, DeepLION introduced a deep learning framework with alternative convolution filters and 1-max pooling operations to handle the amino acid fragments with different lengths. Then, the multi-instance learning framework modeled the TCR correlations and assigned adjusted weights for each TCR sequence during the predicting process. To validate the performance of DeepLION, we conducted a series of experiments on several cohorts of patients from nine cancer types. Compared to the existing methods, DeepLION achieved, on most of the cohorts, higher prediction accuracies, sensitivities, specificities, and areas under the curve (AUCs), where the AUC reached notably 0.97 and 0.90 for thyroid and lung cancer cohorts, respectively. Thus, DeepLION may further support the detection of cancers from TCR repertoire data. DeepLION is publicly available on GitHub, at https://github.com/Bioinformatics7181/DeepLION, for academic usage only.
Collapse
Affiliation(s)
- Ying Xu
- Department of Computer Science and Technology, School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an, China
- Institute of Data Science and Information Quality, Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Xinyang Qian
- Department of Computer Science and Technology, School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an, China
- Institute of Data Science and Information Quality, Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Xuanping Zhang
- Department of Computer Science and Technology, School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an, China
- Institute of Data Science and Information Quality, Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Xin Lai
- Department of Computer Science and Technology, School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an, China
- Institute of Data Science and Information Quality, Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Yuqian Liu
- Department of Computer Science and Technology, School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an, China
- Institute of Data Science and Information Quality, Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
| | - Jiayin Wang
- Department of Computer Science and Technology, School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an, China
- Institute of Data Science and Information Quality, Shaanxi Engineering Research Center of Medical and Health Big Data, Xi’an Jiaotong University, Xi’an, China
- *Correspondence: Jiayin Wang,
| |
Collapse
|
4
|
Wang W, Shen C, Zhao Y, Sun B, Bai N, Li X. Identification and validation of potential novel biomarkers to predict distant metastasis in differentiated thyroid cancer. Ann Transl Med 2021; 9:1053. [PMID: 34422965 PMCID: PMC8339873 DOI: 10.21037/atm-21-383] [Citation(s) in RCA: 8] [Impact Index Per Article: 2.7] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 01/25/2021] [Accepted: 05/16/2021] [Indexed: 12/18/2022]
Abstract
Background Distant metastasis (DM) is not common in differentiated thyroid cancer (DTC). However, it is associated with a significantly poor prognosis. Early detection of high-risk DTC patients is difficult, and the molecular mechanism is still unclear. Therefore, the present study aims to establish a novel predictive model based on clinicopathological parameters and DM-related gene signatures to provide guidelines for clinicians in decision making. Methods Weighted gene co-expression network analysis (WGCNA) was performed to discover co-expressed gene modules and hub genes associated with DM. Univariate and multivariate analyses were carried out to identify independent clinicopathological risk factors based on The Cancer Genome Atlas (TCGA) database. An integrated nomogram prediction model was established. Finally, real hub genes were validated using the GSE60542 database and various thyroid cell lines. Results The midnightblue module was most significantly positively correlated with DM (R=0.56, P=9e-06) by as per WGCNA. DLX5 (AUC: 0.769), COX6B2 (AUC: 0.764), and LYPD1 (AUC: 0.760) were determined to be the real hub genes that play a crucial role in predicting DM. Meanwhile, univariate and multivariate analyses demonstrated that T-stage (OR, 15.03; 95% CI, 1.75-319.40; and P=0.024), histologic subtype (OR, 0.17; 95% CI, 0.03-0.92; and P=0.042) were the independent predictors of DM. Subsequently, a nomogram model was constructed based on gene signatures and independent clinical risk factors exhibited good performance. Additionally, the mRNA expressions of real hub genes in the GSE60542 dataset were consistent with TCGA. Conclusions The present study has provided a reliable model to predict DM in patients with DTC. This model is likely to serve as an individual risk assessment tool in therapeutic decision-making.
Collapse
Affiliation(s)
- Wenlong Wang
- Thyroid Surgery Department, Xiangya Hospital, Central South University, Changsha, China.,National Clinical Research Center for Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China
| | - Cong Shen
- Thyroid Surgery Department, Xiangya Hospital, Central South University, Changsha, China
| | - Yunzhe Zhao
- Thyroid Surgery Department, Xiangya Hospital, Central South University, Changsha, China
| | - Botao Sun
- Thyroid Surgery Department, Xiangya Hospital, Central South University, Changsha, China
| | - Ning Bai
- Thyroid Surgery Department, Xiangya Hospital, Central South University, Changsha, China
| | - Xinying Li
- Thyroid Surgery Department, Xiangya Hospital, Central South University, Changsha, China.,National Clinical Research Center for Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China
| |
Collapse
|