Identification of biosignatures for estimating the aggressiveness of pediatric liver cancer cells from single-cell RNA sequencing

Keng-Shih Chang (張耕石) 1, Tsai-Hsun Tu2, Yi-Tzu Lo1, Yen-Ping Yeh1, Hua-Sheng Chiu3, Pavel Sumazin3, Ting-Wen Chen1,4,5

1Institute of Bioinformatics and Systems Biology, National Yang Ming Chiao Tung University

2Institute of Molecular Medicine and Bioengineering, National Yang Ming Chiao Tung University

3Texas Children’s Cancer Center, Baylor College of Medicine

4Department of Biological Science and Technology, National Yang Ming Chiao Tung University

5Center For Intelligent Drug Systems and Smart Bio-devices (IDS2B), National Yang Ming Chiao Tung University

Pediatric liver cancer is a rare liver cancer and can be divided into two major histological subtypes, hepatoblastoma (HB) and hepatocellular carcinoma (HCC). Usually, HB patients response well to surgical resection and chemotherapy and have a better prognosis compared to HCC patients. However, some pediatric liver cancer cannot be histologically classified as HB or HCC. Therefore, we want to identify biosignatures that can discriminate aggressive pediatric liver cancer cells from nonaggressive ones. With single-cell RNA sequencing data for seven pediatric liver cancer samples from Texas Children’s Cancer Center, we developed a pathway-based scoring system for cancer cells. We identified several previously reported liver cancer biomarkers, including IGF2, DUSP5, HNF1A and HNF4A, as well as pathways such as hepatocellular carcinoma, Wnt signaling pathway, Hippo signaling pathway and fatty acid metabolism. We further validated the pathway-based score with patient-derived xenograft (PDX) models. We created  a tree diagram based on the correlation of pathway scores, which allowed us to separate the cells into subgroups of sensitive and resistant. The results of PDX were consistent with this of our our pathway scoring system. Our study provides a process for researching pediatric liver cancer. In summary, by calculating the pathway-based score, we can separate the cells into aggressive and nonaggressive. It was a significant contribution to the treatment of pediatric liver cancer. Based on our methods, future research can focus on high-risk cells to find more effective treatments.