Related Experiment Video
Updated: Jun 16, 2025
10:20
Single-cell Profiling of Developing and Mature Retinal Neurons
Published on: April 19, 2012
14.1K
Integrating explainable machine learning and transcriptomics data reveals cell-type specific immune signatures
Khang Ma1, Hosei Nakajima1, Nipa Basak1
1Department of Ophthalmology, Baylor College of Medicine, Houston, TX, USA.
NPJ Genomic Medicine
|June 14, 2025
Summary
Immune cells, particularly microglia and astrocytes in the retina, play a crucial role in Age-related Macular Degeneration (AMD). This study identifies key genes and a novel genetic variant linked to AMD progression.
Area of Science:
- Ophthalmology
- Immunology
- Genetics
Background:
- Genome-wide association studies (GWAS) implicate immune dysfunction in Age-related Macular Degeneration (AMD).
- The specific roles of immune cells in AMD pathogenesis remain incompletely understood.
Purpose of the Study:
- To identify genes and cellular mechanisms underlying AMD using an explainable machine learning approach.
- To investigate the contribution of retinal glial cells, specifically microglia and astrocytes, to AMD.
Main Methods:
- Developed an explainable machine learning pipeline utilizing transcriptome data from 453 donor retinas.
- Employed cellular deconvolution and analyzed independent single-cell RNA sequencing data.
- Integrated AMD-GWAS data to identify regulatory variants.
Main Results:
- Identified 81 genes distinguishing AMD from controls with high accuracy (AUC-ROC 0.80).
- Found enrichment of these genes in retinal glial cells (microglia and astrocytes).
- Detected significant differences in microglia and astrocytes between normal and AMD retinas and identified a novel AMD-associated variant (rs4133124 at PLCG2).
Conclusions:
- Retinal glial cells, particularly microglia and astrocytes, are significantly implicated in AMD.
- The study provides molecular insights into immune dysfunction in AMD and highlights glial cell heterogeneity in disease progression.
