Related Experiment Video
Updated: Jul 14, 2026
09:24
3-D Cell Culture System for Studying Invasion and Evaluating Therapeutics in Bladder Cancer
Published on: September 13, 2018
9.5K
Integrating transcriptomics, single-cell omics, and deep learning-based histopathological features to identify OLFML3
Fazhong Dai1, Yifeng He1,2, Xiongsheng Huang1
1Department of Urology, The Affiliated Guangdong Second Provincial General Hospital of Jinan University, Guangzhou, China.
Translational Andrology and Urology
|November 13, 2025
Summary
This study identifies OLFML3 as a key gene linked to bladder cancer (BCa) recurrence within one year. A deep learning model predicts OLFML3 expression from pathology images, aiding in BCa prognosis.
Area of Science:
- Oncology
- Genomics
- Bioinformatics
Background:
- Bladder cancer (BCa) has a high recurrence rate (approx. 61% within 1 year post-surgery).
- Current monitoring methods (cystoscopy, urine cytology) have low sensitivity and cause patient discomfort.
- There is a need for improved accuracy in predicting BCa recurrence.
Purpose of the Study:
- To investigate the role of OLFML3 in BCa recurrence using multi-omics data.
- To enhance the accuracy of BCa recurrence prediction.
- To improve clinical management strategies for BCa.
Main Methods:
- RNA sequencing (RNA-seq) and clinical data from TCGA were analyzed.
- Weighted gene co-expression network analysis (WGCNA) identified recurrence-associated gene modules.
- Univariate and LASSO Cox regression selected eight prognostic genes.
- Single-cell RNA sequencing (scRNA-seq) evaluated gene expression in BCa subtypes.
- A ResNet50-based deep learning model predicted OLFML3 expression from H&E images.
Main Results:
- WGCNA identified a red module positively correlated with BCa recurrence.
- Eight prognostic genes were selected, with high OLFML3 expression linked to lower survival rates.
- scRNA-seq showed differential OLFML3 expression in BCa subtypes.
- High OLFML3 expression correlated with BCa invasiveness, grade, and stage.
- A random forest model accurately predicted OLFML3 expression levels from pathological features.
Conclusions:
- Multi-omics analysis identified OLFML3 as a critical factor in 1-year BCa recurrence.
- A deep learning model can predict OLFML3 expression from pathological images.
- Further research is needed to fully elucidate OLFML3's role in BCa recurrence.
