Accurate Microsatellite Instability Classification Using Ultima Genomics Platform Across Multiple Tumor Types

Microsatellite instability (MSI), a hallmark of mismatch repair deficiency (MMRd), occurs across multiple cancer types and carries distinct diagnostic, prognostic, and therapeutic implications. Accurate MSI detection is therefore essential in clinical oncology. Traditional approaches - mismatch repair Immunohistochemistry (IHC) and Polymerase Chain Reaction (PCR)-based assays - are widely used but may misclassify tumors, particularly in cancer types with a low prevalence of MSI. Although next-generation sequencing (NGS)-based methods provide improved accuracy, their adoption is constrained by cost. Here, we evaluate the UG 100 ^® , a low-cost flow-based sequencing platform, for its ability to classify MSI status across diverse tissues and sequencing conditions. We developed and assessed complementary computational strategies, including (I) tumor–normal microsatellite indel (MS-indel) calling and (II) per-read and homopolymer variant analyses, supported by tumor-purity simulation experiments to test robustness in clinically realistic scenarios. Across a diverse validation cohort spanning cell lines, Fresh-Frozen (FF), and Formalin-Fixed Paraffin-Embedded (FFPE) tumors, each method achieved near-perfect Area Under the ROC curve (AUROC). In dilution experiments, the MSI-score remained informative at low tumor fractions, distinguishing MSI from MSS cell lines down to ~ 5%. These findings demonstrate that UG 100 ^® platform enables reliable MSI detection from low-coverage, tumor-only sequencing data. This capability positions flow-based NGS as a practical and cost-effective alternative to conventional MSI assays, with strong potential for broad clinical implementation.