MAPS

MAPS applies high-content microscopy and deep learning to assess functional effects of genetic variants by detecting changes in protein subcellular localization.

Key Features:

• High-Content Microscopy Integration: Analyzes high-content microscopy (HCM) images to detect alterations in protein subcellular localization as indicators of loss of function.
• Deep Learning Analysis: Uses deep learning algorithms to analyze large-scale HCM imaging datasets.
• Cloud-Based Deep Learning: Utilizes cloud-based deep learning services for model training and deployment.
• Variant Functionalization: Demonstrated application to functionalize PTEN missense variants by assessing their impact on protein localization.

Scientific Applications:

• Variant Classification in Cancer Genetics: Functionalizes and aids classification of variants of uncertain significance in hereditary and somatic cancers by evaluating localization changes.
• Tumor Suppressor Assessment: Evaluates tumor suppressor genes such as PTEN where subcellular localization informs loss-of-function status.

Methodology:

Analysis of high-content microscopy images using deep learning algorithms with model training and deployment performed via cloud-based deep learning services.