Thank you to everyone involved! Maria Capilla-Guerra, Andre Monteleone, Jonathan Young, Subha Shankar, Phillip Michaels, Harrison Tsai, @jdgood1.bsky.social, Antonia Kreso, @vangalenlab.bsky.social, Christoph Schliemann and Evan Chen. @danafarber.bsky.social @broadinstitute.org (11/11)

➡️ Moving forward, we are excited to continue developing this framework by expanding to other hematological malignancies, optimizing MARLIN’s integration into existing diagnostic workflows, and validating MARLIN’s accuracy and impact in broader prospective studies. (10/n)

- When coupled with nanopore sequencing, MARLIN provides accurate real-time classifications, often within two hours from sample receipt. (9/n)

- MARLIN is a neural network for DNA methylation-based acute leukemia classification that shows high accuracy in retrospective and prospective samples. (8/n)

…but also reveal novel disease heterogeneity or shared biology that is not captured by conventional testing alone. (7/n)

- Methylation classes frequently mirror established molecular categories in acute leukemia…(6/n)

- DNA methylation profiling is a valuable approach for assigning acute leukemia lineages and it may help classify lineage ambiguous acute leukemia. (5/n)

🔬Key takeaways:
- We built a comprehensive reference of pediatric + adult acute leukemia and defined 38 acute leukemia methylation classes. (4/n)

💡To address these issues, we have developed a comprehensive framework for rapid classification of acute leukemia using epigenetics, nanopore sequencing and a novel machine learning classifier for real-time acute leukemia classification. (3/n)

❓Diagnosing acute leukemia quickly and accurately is crucial for its treatment, but current methods are time-consuming and do not necessarily identify all features of the disease. (2/n)