MSI Expresso: a software for determining MSI status and detecting MSI-related transcription events from RNA sequencing data

Emmanuel Tubacher^1*Alexandre How-Kit¹Mourad Sahbatou¹Alex Duval²Victor RENAULT¹JF Deleuze¹

• ¹Fondation Jean Dausset Centre d'Etude du Polymorphisme Humain, Paris, France
• ²Centre de recherche Saint-Antoine, Inserm UMR_S 938, équipe "instabilité des microsatellites et cancers", 75012 Paris, France., Paris, France

Microsatellite instability (MSI) is becoming increasingly important in oncology as it has been reported across more than two dozen of solid cancer types. The MSI-high phenotype has long been used as a predictive and prognostic marker in colorectal cancer and has been recently approved by the FDA as a marker for immune checkpoint blockade therapy for solid cancers. Several bioinformatics tools have been developed to assess MSI status of a tumor sample using Next-Generation Sequencing (NGS) data mostly from whole genome, whole exome, and targeted gene sequencing data. While most tools available only infer the MSI status, none of them use RNA-sequencing (RNA-seq) data and provide per microsatellite expression and genotype results. We present MSI Expresso, a software which assesses the MSI status by testing the instability of a panel of 3'UTR microsatellites from RNA-seq data and also provides a detailed landscape of MSI-related events such as exon skipping, unstable coding and intronic microsatellites with a graphical output of the recurrent events. MSI Expresso's ability to detect the MSI status was assessed from RNA-seq data of 228 colon, 13 prostate and 2 endometrial cancer samples with known MSI status and achieved almost 100% concordant results. Thus, MSI Expresso is a new tool for MSI detection from RNA-sequencing data complementary to genomic and genetic approaches allowing to explore the consequence of MSI events on transcripts/transcriptome.