Method for constructing prognostic model of hepatoma and application thereof

Patent metadata

Field	Value
Publication number	US-12469607-B2
Application number	US-202519035563-A
Country	US
Kind code	B2
Filing date	Jan 23, 2025
Priority date	Mar 25, 2024
Publication date	Nov 11, 2025
Grant date	Nov 11, 2025

Abstract

The disclosure belongs to the field of genetic testing and biomedicine, relating to a method for constructing a prognostic model of hepatoma and an application thereof, comprising 1) obtaining and identifying fibroblasts with high FAP expression; 2) obtaining and identifying TAMs; 3) analyzing co-localization between fibroblasts with high FAP expression obtained and the TAMs obtained previously; 4) communicating and analyzing the fibroblasts with high FAP expression after the localization in the Step 3) with TAMs to obtain CCC ligand-receptor genes; 5) screening the CCC ligand-receptor genes obtained previously based on machine learning to obtain key CCC ligand-receptor genes; and 6) constructing a prognostic model of hepatoma according to the key CCC ligand-receptor genes obtained in the Step 5). The present disclosure provides a method for constructing a prognostic model of hepatoma that can be applied to auxiliary judgment of the prognosis of hepatoma patients and an application thereof.

First claim

What is claimed is:

1 . A method for constructing a prognostic model of hepatoma, the method comprising the following steps:
(1) obtaining and identifying fibroblasts with high FAP expression, specifically as follows: (1.1) performing a subgroup classification of hepatoma single-cell data in a collected and integrated discovery cohort using R package seurat to extract fibroblast subgroups with high COL1A1 expression; and (1.2) further subdividing the fibroblast subgroups with high COL1A1 expression obtained in the Step (1.1) to identify fibroblasts with high FAP expression; (2) obtaining and identifying tumor-associated macrophages (TAMs), specifically as follows; (2.1) performing a subgroup classification of hepatoma single-cell data in a collected and integrated discovery cohort using R package seurat to extract macrophage subgroups with high CD68 expression; (2.2) further subdividing the extracted macrophage subgroups with high CD68 expression; (2.3) performing an OR analysis to assess the enrichment preference of different cell types in different samples and screening for cell types highly enriched in hepatoma samples; (2.4) reconstructing a macrophage differentiation process using an RNA rate analysis in another collected and integrated single-cell validation cohort, and identifying the type of macrophages that terminally differentiate with tumor development as TAMs; the macrophage type being high Disabled-2 (DAB2) expression or high Secreted Phosphoprotein 1 (SPP1) expression; and (3) co-localizing the fibroblasts with high FAP expression obtained in the Step (1) and the TAMs obtained in the Step (2); (4) communicating and analysing the fibroblasts with high FAP expression after the localization in the Step (3) with TAMs to obtain CCC ligand-receptor genes, specifically as follows; (4.1) identifying the CCC ligand-receptors between TAMs and fibroblasts with high FAP expression using R package NicheNet, and identifying the target genes of fibroblasts with high FAP expression affected by TAMs; (4.2) analysing, the function of target genes by g: Profiler to understand the main functional regulation of TAMs on fibroblasts with high FAP expression; (4.3) scoring a tissue sequencing sample based on the activity of CCC ligand-receptors using the ssGSEA algorithm, and the scoring result being LRscore; (4.4) identifying a cutoff value of optimal survival probability grouping of samples by R package survminer, and testing the predictive effect of LRscore on an overall survival probability of patients by Kaplan-Meier curves, wherein if log-rank p<0.05, the test standards are satisfied; (4.5) on the basis of the Step (4.4), testing the predictive effect of LRscore on immunotherapy response in patients with hepatoma by box plots, wherein if wilcox.testp<0.05, the test standards are satisfied; and (4.6) obtaining CCC ligand-receptor genes based on the result of the Step (4.5); (5) screening the CCC ligand-receptor genes obtained in the Step (4) based on machine learning to obtain key CCC ligand-receptor genes, the key CCC ligand-receptor genes are CD320, GPC1, ITGA5 and ENG;
and;
(6) constructing a prognostic model of hepatoma according to the key CCC ligand-receptor genes obtained in the Step (5);
(6.1) based on the modeling genes determined in the Step (5), constructing a multivariate Cox model in the TCGA and GEO hepatoma cohorts, calculating model scores, and the model score being calculated according to the following equation:
Coxmodel score=Σ i Expression (mRNA) i *Coefficent (mRNA) i

where i is the key gene screened;
(6.2) using KM curves to evaluate the survival prediction performance of the model constructed in the Step (6.1);
(6.3) predicting patients' response to immunotherapy based on Coxmodel score, wherein non-responders exhibit higher Coxmodel scores than responders, and patients are classified into a non-responder group and a responder group according to the Coxmodel scores, and the method further comprises:
administering a first treatment regimen comprising an immunotherapy to a patient classified into the responder group.

2 . The method according to claim 1 , wherein the specific implementation manner of the Step (3) is as follows: mapping the identified single-cell subgroup to spatial transcriptome sequencing sections using R package CellTrek, and confirming that fibroblasts with high FAP expression have a high spatial proximity to TAMs by Kullback-Leibler divergence.

3 . The method according to claim 2 , wherein the specific implementation manner of the Step (5) is as follows:
(5.1) in the TCGA HCC cohort, genes with log-rank p<0.05 being further screened from those constructed for LRscoring using univariate Cox analysis; and (5.2) using such machine learning algorithms as Stepcox, RSF, LASSO and CoxBoost to determine key genes from the genes screened in the Step (5.1), respectively, and defining an intersection of key genes to obtain final modeling genes.

4 . A prognostic model of hepatoma obtained by the method according to claim 1 .

5 . An application of the prognostic model of hepatoma according to claim 4 in auxiliary judgment of disease prognosis.

6 . An application of the prognostic model of hepatoma according to claim 4 in auxiliary judgment of hepatoma prognosis.

Assignees

• Univ Wuhan

Inventors

• Wang Fubing
• Long Fei
• Zhong Wei

Classifications

• G16H50/20

  for computer-aided diagnosis, e.g. based on medical expert systems · CPC title

• G16B40/20

  Supervised data analysis · CPC title

• G16B20/00

  ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations · CPC title

• G16B5/20 Primary

  Probabilistic models · CPC title

• G16H20/10

  relating to drugs or medications, e.g. for ensuring correct administration to patients · CPC title