Back

M&M: An RNA-seq based Pan-Cancer Classifier for Pediatric Malignancies

Wallis, F. S. A.; Baker-Hernandez, J. L.; van Tuil, M.; van Hamersveld, C. A.; Koudijs, M. J.; Verwiel, E. T. P.; Janse, A.; Hiemcke-Jiwa, L. S.; de Krijger, R. R.; Kranendonk, M. E. G.; Vermeulen, M. A.; Wesseling, P.; Flucke, U. E.; de Haas, V.; Luesink, M.; Hoving, E. W.; Vormoor, H. J.; van Noesel, M. M.; Hehir-Kwa, J. Y.; Tops, B. B. J.; Kemmeren, P.; Kester, L. A.

2024-06-07 oncology
10.1101/2024.06.06.24308366 medRxiv
Show abstract

With many rare tumor types, acquiring the correct diagnosis is a challenging but crucial process in pediatric oncology. Here, we present M&M, a pan-cancer ensemble-based machine learning algorithm tailored towards inclusion of rare tumor types. The RNA-seq based algorithm can classify 52 different tumor types (precision[~] 99%, recall[~] 80%), plus the underlying 96 tumor subtypes (precision[~] 96%, recall[~] 70%). For low-confidence classifications, a comparable precision is achieved when including the three highest-scoring labels. M&Ms pan-cancer setup allows for easy clinical implementation, requiring only one classifier for all incoming diagnostic samples, including samples from different tumor stages and treatment statuses. Simultaneously, its performance is comparable to existing tumor- and tissue-specific classifiers. The introduction of an extensive pan-cancer classifier in diagnostics has the potential to increase diagnostic accuracy for many pediatric cancer cases, thereby contributing towards optimal patient survival and quality of life.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.