SynLeGG: analysis and visualization of multiomics data for discovery of cancer 'Achilles Heels' and gene function...

Area of Science:

• Genomics
• Computational Biology
• Cancer Research

Background:

• Achilles' heel relationships, where one gene's status sensitizes cells to another's perturbation, offer therapeutic targets in precision oncology.
• Investigating genetic dependency relationships (GDRs) using transcriptome data is underexplored despite its advantages over genetic approaches.

Purpose of the Study:

• To introduce SynLeGG, a tool for identifying and visualizing mutually exclusive loss signatures in omics data to discover GDRs.
• To leverage transcriptome data for uncovering novel GDRs and tissue-specific vulnerabilities.

Main Methods:

• SynLeGG utilizes the MultiSEp algorithm for unsupervised clustering of cell lines based on gene expression.
• Analysis involves correlating CRISPR scores and mutational status within identified clusters to propose candidate GDRs.
• Benchmarking against SynLethDB assessed MultiSEp's performance using Receiver Operator Characteristic curves and coverage.

Main Results:

• MultiSEp demonstrated superior performance compared to competing methods, achieving significantly higher area under the ROC curve and 2.8- to 8.5-fold greater coverage.
• SynLeGG successfully identified pan-cancer and tissue-specific GDRs, including the known synthetic lethality of SMARCA2 with SMARCA4.
• Predictions from SynLeGG were significantly enriched in dependencies validated by recent CRISPR screens.

Conclusions:

• SynLeGG provides a robust platform for discovering genetic dependency relationships using transcriptome data.
• The tool aids in prioritizing potential drug targets by integrating multi-omics information and identifying vulnerabilities exploitable in cancer therapy.