Identification of a novel gene signature for the prediction of recurrence in HCC patients by machine learning of genome-wide databases

Abstract

Hepatocellular carcinoma (HCC) is a common malignant tumor in China. In the present study, we aimed to construct and verify a prediction model of recurrence in HCC patients using databases (TCGA, AMC and Inserm) and machine learning methods and obtain the gene signature that could predict early relapse of HCC. Statistical methods, such as feature selection, survival analysis and Chi-Square test in R software, were used to analyze and select mutant genes related to disease free survival (DFS), race and vascular invasion. In addition, whole-exome sequencing was performed on 10 HCC patients recruited from our center, and the sequencing results were compared with the databases. Using the databases and machine learning methods, the prediction model of recurrence was constructed and optimized, and the selected mutant genes were verified in the test group. The accuracy of prediction was 74.19%. Moreover, these 10 patients from our center were used to verify these mutant genes and the prediction model, and a success rate of 80% was achieved. Collectively, we discovered recurrence-related genes and established recurrence prediction model of recurrence for HCC patients, which could provide significant guidance for clinical prediction of recurrence. Hepatocellular carcinoma (HCC) is a common malignant tumor in China, which ranks the third in morbidity and the second in mortality. Its morbidity is usually associated with specific risk factors, including infections with HBV and HCV, high alcohol intake, obesity and consumption of aflatoxin-containing food 1. With the development of the second-generation sequencing techniques increasing research on HCC has been conducted on the molecular level. In 2014, Totoki et al. 2 have reported the whole-genome sequencing of 608 HCC patients from Asia and Europe. In 2015, Schulze et al. 3 have reported the whole-genome sequencing of 243 HCC patients from Europe and America. In 2016, Fujimoto et al. 4 have reported the whole-genome sequencing of 300 HCC patients from Japan. The molecular blueprint of HCC including somatic mutation, mRNA expression, methylation and miRNA regulation has been gradually outlined, which could be used for the diagnosis, treatment, and prediction of recurrence and survival of liver cancer patients. In 2017, TCGA working group 5 has systematically analyzed the sequencing results of the whole exome of more than 360 HCC patients in TCGA database and compared these data with other published HCC sequencing samples. Various statistical methods, related classification and clustering algorithms of machine learning have been used. TERT, TP53, CTNNB1, AXIN1, ARID1A, ARID2, RB1, ALB, APOB, PTEN, CDKN2A, DOCK2 6-15 and other somatic cells with significantly mutant genes (SMGs) and driver mutation have been identified. These findings have been rapidly applied as potential therapeutic targets and prognostic indicators in clinical practice. However, the high cost of whole-exome sequencing and whole-genome sequencing limits its use in clinical practice. Actually, patients often ...

Predictive Modeling of Survival and Toxicity in Patients With Hepatocellular Carcinoma After Radiotherapy

“…These models suffered from small sample size or relied on data not readily available, impeding their clinical application. For instance, Shen et al 42 predicted disease-free survival from gene sequencing but the model was validated on 10 patients only. Others also used powerful deep learning techniques 43 or focused on radiomics, 44 but lacked external validation.…”

Analysis of the TCGA Dataset Reveals that Subsites of Laryngeal Squamous Cell Carcinoma Are Molecularly Distinct

“…Conversely, we found DNAH5, which encodes a protein involved in the dynein complex and functions in ciliary cell motility to have more tobacco associated G>T transversions in the supraglottic cancers [ 35 ]. Mutations in DNAH5 have been associated with poorer survival in a number of different cancers [ 36 , 37 , 38 , 39 ]. We did not note an impact of DNAH5 G>T mutation on survival in this study, but this may reflect the limited sample size of the LSCC cohort.…”

Application of Proteomics in Cancer: Recent Trends and Approaches for Biomarkers Discovery

“…AI can be implemented to create algorithms that increase their performance when certain types of resources or data are provided ( 129 ). Several studies have used AI tools to identify novel cancer biomarkers or predict cancer stages ( 130 ). Shen et al used the Boruta algorithm to identify mutant genes involved in vascular invasion from TCGA, National Institute of Health, Medical Research, and AMC databases.…”