Related Experiment Video
Updated: Sep 22, 2025
07:46
Author Spotlight: Advancing Reproductive Immunology with a Protocol for the Quantitative Evaluation of Endometrial Immune Cells
Published on: October 13, 2023
1.4K
Gradient Boosting Machine Learning Model for Defective Endometrial Receptivity Prediction by Macrophage-Endometrium
Bohan Li1, Hua Duan1, Sha Wang1
1Department of Minimally Invasive Gynecologic Center, Beijing Obstetrics and Gynecology Hospital, Capital Medical University, Beijing Maternal and Child Health Care Hospital, Beijing, China.
Frontiers in Immunology
|May 23, 2022
Summary
Defective endometrial receptivity impacts fertility. This study found that macrophage infiltration levels, analyzed via gene modules, accurately predict reproductive success, outperforming traditional methods like endometrial thickness measurement.
Area of Science:
- Reproductive Biology
- Immunology
- Bioinformatics
Background:
- Defective endometrial receptivity contributes to approximately one-third of infertility cases and implantation failures.
- Investigating immune cell infiltration, particularly macrophages, is crucial for understanding reproductive outcomes.
Purpose of the Study:
- To analyze the association between immune cell infiltration levels and reproductive outcomes.
- To develop an accurate and cost-effective method for assessing endometrial receptivity using macrophage-endometrium interactions.
Main Methods:
- Pooled analysis of 218 subjects from multiple datasets (GEO).
- Construction of macrophage-endometrium interaction modules using weighted gene co-expression network and differential gene expression analysis.
- Development and validation of predictive models using machine learning algorithms (Xgboost, random forests, regression) and clinical samples.
Main Results:
- Altered macrophage (Mϕ) infiltration levels significantly influence embryo implantation.
- Selected gene modules highlight macrophage-endometrium interactions involved in immunoreactivity, decidualization, and signaling.
- The Xgboost model demonstrated superior predictive performance (AUCs up to 0.998) compared to other models and ultrasonography for endometrial thickness.
Conclusions:
- Macrophage infiltration levels, reflected in specific genetic modules, are key indicators of successful embryo implantation.
- Hub genes within these modules offer a basis for developing advanced machine learning models to predict reproductive outcomes in women with compromised endometrial receptivity.
