网络增强核融合方法的改进及其在乳头状肾细胞癌多组学数据整合分子分型中的应用

An Improved Network-Enhanced Fusion Method and its Application in Papillary Renal Cell Carcinoma Subtyping using Multi-omics Data

目的针对网络增强的相似网络融合(network enhancement fusion,ne-SNF)方法先融合不同组学网络,再对融合后的网络降噪,忽略了不同组学相似网络噪声对融合网络影响的问题,本文提出了改进的网络增强融合(improved network enhancement fusion,improved ne-SNF)方法,并探讨其在乳头状肾细胞癌(papillary renal cell carcinoma,PRCC)分子分型中的应用,识别PRCC高危患者,筛选重要通路及免疫浸润细胞。方法通过模拟研究评估improved ne-SNF分型性能,并将其用于PRCC多组学数据的整合分型,利用Cox回归模型分析不同分型患者的预后风险;筛选不同分型的差异表达mRNA(DEmRNAs)、miRNA(DEmiRNAs)及差异甲基化基因(DMGs),并对其重合基因进行KEGG通路分析;最后对不同分型患者进行免疫细胞浸润分析。结果模拟研究结果表明improved ne-SNF在不同信号比例和噪声强度下的分型准确性均优于SNF和ne-SNF。improved ne-SNF方法将PRCC患者分为高危组和低危组,高危组患者的死亡风险是低危组的7.727倍;筛选出3511个DEmRNAs,96个DEmiRNAs及3426个DMGs,其联合分析的649个重合基因得到42条有统计学差异的KEGG通路。此外,筛选出3种在不同分型中存在统计学差异的免疫浸润细胞。结论improved ne-SNF分型性能优于SNF和ne-SNF,且能够有效识别PRCC预后高风险患者,并筛选出PRCC重要通路及相关免疫浸润细胞,为PRCC的治疗及预后提供新的思路和参考依据。

Objective The Network Enhancement Fusion(ne-SNF)method had no denoising strategy for the networks from each omics data type,which would weaken the power of network fusion.To address this problem,we proposed an improved Network Enhancement Fusion(improved ne-SNF)model,and further applied the method to subtype identification of papillary renal cell carcinoma.Methods We conducted simulation studies to compare the performances of the improved ne-SNF method with the ne-SNF and SNF method,and applied improved ne-SNF method to integrate multi-omics data of PRCC.Cox regression model was performed to evaluate the prognostic risk of different subtypes.Differentially expressed mRNAs(DEmRNAs),miRNAs(DEmiRNAs)and differentially methylated genes(DMGs)with different subtypes were screened.KEGG pathway analysis was performed for the overlapping genes of three gene sets.Finally,the immune cell infiltration analysis was performed for patients with different subtypes.Results The improved ne-SNF method outperformed both SNF and ne-SNF approach in various simulation scenarios.In subsequent subtyping application,PRCC patients were divided into high-risk and low-risk groups,and the risk of death was 7.727 times higher in the high-risk group than in the low-risk group.A total of 3511 DEmRNAs,96 DEmiRNAs and 3426 DMGs were identified.Among them,649 overlapping genes yielded 42 KEGG pathways with statistical differences.In addition,3 immune filtrating cells showed statistical significance.Conclusion The improved ne-SNF performed better than SNF and ne-SNF,and the identified subtypes of PRCC may provide important clues and basis for treatment of PRCC patient.