Image-based, pooled phenotyping reveals multidimensional, disease-specific variant effects.

Area of Science:

• Genomics
• Cell Biology
• Bioinformatics

Background:

• Genetic variants often have complex phenotypic effects that are difficult to assay.
• Current predictive models struggle to capture the full spectrum of variant impacts.
• Understanding variant function is crucial for diagnosing genetic diseases.

Purpose of the Study:

• To develop a novel method for measuring variant effects on molecular and cellular phenotypes.
• To apply this method to a large set of genetic variants in LMNA and PTEN genes.
• To establish a framework for resolving the complexity of genetic variant function.

Main Methods:

• Developed Variant in situ sequencing (VIS-seq), a pooled, image-based assay.
• Applied VIS-seq to approximately 3,000 LMNA and PTEN variants across diverse cell types.
• Generated high-dimensional morphological profiles to capture variant-driven changes.

Main Results:

• VIS-seq identified variant-specific changes in protein abundance, localization, activity, and cell architecture.
• Discovered gain-of-function LMNA variants that alter nuclear shape and autism-associated PTEN variants with mislocalization.
• Morphological profiles achieved near-perfect accuracy in predicting variant pathogenicity and distinguishing disease-associated variants.

Conclusions:

• Most genetic variants affect a multidimensional continuum of phenotypes, not captured by single assays.
• VIS-seq provides a scalable framework for linking genetic variation to cellular phenotypes.
• This approach illuminates the cascade of variant effects from molecular to morphological levels, aiding in understanding genetic disease complexity.