AUTHOR=Qin Da , Guo Qingdong , Wei Rui , Liu Si , Zhu Shengtao , Zhang Shutian , Min Li 

TITLE=Predict Colon Cancer by Pairing Plasma miRNAs: Establishment of a Normalizer-Free, Cross-Platform Model

JOURNAL=Frontiers in Oncology

VOLUME=Volume 11 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2021.561763

DOI=10.3389/fonc.2021.561763

ISSN=2234-943X

ABSTRACT=Background: Plasma miRNAs are emerging biomarkers for colon cancer (CC) diagnosis. However, the lack of robust internal references largely limited their clinical application. Here we proposed a miRNA ratio-based, normalizer-free algorithm to quantitate plasma miRNA.
Methods: A miRNA-miRNA pair matrix was established by pairing DE miRNAs in training group from GSE106817. LASSO regression was performed to build prediction models. To maximize the performance of prediction models, four algorithms (lasso regression, random forest, logistic regression, and SVM) were simultaneously performed for each biomarker combination. Chip data from testing group from GSE106817 and GSE112264 were used for internal and external verification, separately. RT-qPCR data from 104 plasma samples from hospital were also used for external validation. 
Results: After validation via 4 different methods, we found 4-pair model could well differentiate CC patients from normal controls with a minimal AUC of 0.93 in internal verification (GSE106817) and 0.78 in external verification (GSE112264). Further validation using RT-qPCR data exhibited its good classifier ability with an AUC of 0.78. 
Conclusion: We constructed a cross-platform prediction model robust against sample-specific disturbance, which is not only well-performed in prediction CC but also promising in the diagnosis of other cancers.