MCVAE-based multi-omic anomaly detection in Fragile X Syndrome

Area of Science:

• Neuroscience
• Computational Biology
• Genetics

Background:

• Fragile X Syndrome (FXS) is a genetic neurodevelopmental disorder caused by mutations in the FMR1 gene, leading to the absence of Fragile X Messenger Ribonucleoprotein (FMRP).
• FMRP regulates the translation of numerous mRNAs, and its loss disrupts the coordination between transcriptomic and translatomic data in the FXS brain.
• Existing omics analyses face challenges due to limited sample availability and dataset heterogeneity, hindering the detection of subtle, coordinated multi-omic dysregulations.

Purpose of the Study:

• To develop a computational framework for uncovering coordinated molecular perturbations in FXS.
• To leverage a Multi-Channel Variational Autoencoder (MCVAE) to integrate transcriptomic and translatomic data for improved analysis.
• To identify potential biomarkers and therapeutic targets for FXS by understanding its underlying pathophysiology.

Main Methods:

• Trained a Multi-Channel Variational Autoencoder (MCVAE) on wild-type samples to learn a shared latent representation of transcriptomic and translatomic data through cross-modal reconstruction.
• Applied the trained MCVAE to Fmr1 knock-out samples (modeling FXS) to detect deviations from wild-type patterns as anomalies.
• Utilized publicly available databases to explore coordinated relationships between transcriptomic and translatomic anomalies and their mapping to known biological pathways.

Main Results:

• The MCVAE successfully identified known and novel molecular perturbations in the Fmr1 knock-out model.
• MCVAE demonstrated superior performance compared to alternative methods, showing stronger enrichment for FMRP mRNA targets and enhanced genotype discriminative power.
• Detected coordinated anomalies between transcriptomic and translatomic layers that mapped to validated FMRP regulators and crucial neurodevelopmental pathways.

Conclusions:

• MCVAE serves as a powerful framework for uncovering coordinated molecular dysregulations in complex diseases like FXS.
• The identified anomalies provide insights into the pathophysiology of FXS, highlighting the interplay between gene expression and protein translation.
• This approach facilitates the identification of novel biomarkers and therapeutic targets for Fragile X Syndrome.