Performance evaluation for rapid detection of pan-cancer microsatellite instability with MANTIS

• DOI: 10.18632/oncotarget.13918
• Publication Date: 2016-12-13T18:04:09Z
• Authors: Esko A. Kautto; Russell Bonneville; Jharna Miya; Lianbo Yu; Melanie A. Krook; Julie W. Reeser; Sameek Roychowdhury
• Keywords: Mantis; Microsatellite Instability

// Esko A. Kautto 1 , Russell Bonneville Jharna Miya Lianbo Yu 2 Melanie Krook Julie W. Reeser and Sameek Roychowdhury 1,3 Comprehensive Cancer Center, The Ohio State University, Columbus, OH, USA Department of Biomedical Informatics, 3 Division Medical Oncology, Internal Medicine, Correspondence to: Roychowdhury, email: Keywords : microsatellite instability, computational biology, next-generation sequencing Received November 24, 2016 Accepted December 02, Published 12, Abstract In current clinical practice, instability (MSI) mismatch repair deficiency detection is performed with MSI-PCR immunohistochemistry. Recent research has produced several tools for MSI (NGS) data; however a comprehensive analysis methods not yet been performed. this study, we introduce new tool, MANTIS, demonstrate its favorable performance compared to the previously published mSINGS MSISensor. We evaluated 458 normal-tumor sample pairs across six cancer subtypes, testing classification on variable numbers target loci ranging from 10 2539. All three were found be accurate, MANTIS exhibiting highest accuracy 98.91% samples all diseases classified correctly. displayed superior among tools, having overall sensitivity (MANTIS 97.18%, MSISensor 96.48%, 76.06%) specificity 99.68%, 98.73%) types, even panels varying size. Additionally, also had lowest resource consumption (<1% space <7% memory required by mSINGS) fastest running times (49.6% 8.7% mSINGS, respectively). This study highlights potential utility in classifying MSI-status, allowing incorporation into existing NGS pipelines.