A new algorithm classifies different types of sarcoma

  • An international study has developed a new algorithm that classifies a wide range of sarcomas according to their molecular patterns.
  • Currently, the algorithm classifies around 70% of cases correctly. These good results open a new and more efficient way of diagnosis.
Imatge Nature Tirado 9

Last week, it was published a new algorithm that analyses sarcomas. The algorithm classifies these malignant tumours according to a known molecular fingerprint: the DNA methylation pattern. Three researchers from IDIBELL has participated in the study, published in the Nature Communications journal. At the moment, the algorithm assigns correctly around 70% of sarcomas and it is available at molecularsarcomapathology.org for free.

Sarcomas are malignant tumours of connective tissue, both soft (including muscles, fat or blood vessels) and bone. It is difficult to distinguish between the types of sarcomas due to their high morphological variability. For this reason, assigning a correct prognosis and treatment is not an easy task. Since histopathologists have issues to classify sarcomas, currently they are already diagnosed through molecular techniques. Specifically, biopsies are used to detect possible gene fusions, a kind of DNA mutation that can cause cancer, but it does not explain all cases.

DNA is modulated by the environment and external factors. Corresponding adjustments along the DNA sequence regulate the expression of genes. They are what scientists call epigenetic modifications and contrary to mutations, genes are not modified. One of the best known epigenetic modifications are DNA methylations. The pattern of methylations along the DNA structure is specific of each cellular type, including tumour cells. Each pattern represents a molecular fingerprint that this study has used to develop the sarcomas classifier. Specifically, researchers have trained an algorithm with the methylation patterns of 1077 sarcomas, successfully characterized in advance. After training, the predictive power of the algorithm has been verified in 428 different sarcomas. The algorithm has correctly classified 263 cases (61%) previously diagnosed by former methods. What is more, it has corrected the diagnosis of 29 cases (7%) that had been wrongly classified. Only in 4 cases (1%) the prediction was wrong.

The authors of the study state that the algorithm can be improved. On one hand, they plan to train it with more samples so unusual sarcomas are better represented. On the other hand, they suspect that some samples contain a great proportion of non-tumour cells within the sarcoma, typically inflammatory cells. In this sense they would optimize the algorithm by including the methylation patterns of this other kind of cells.


Image: Adapted from Lena Jakob, Gisela Metzler, Ko-Ming Chen, Claus Garbe / wikicommons and Starline / Freepik


The Bellvitge Biomedical Research Institute (IDIBELL) is a biomedical research center created in 2004. It is participated by the Bellvitge University Hospital and the Viladecans Hospital of the Catalan Institute of Health, the Catalan Institute of Oncology, the University of Barcelona and the City Council of L’Hospitalet de Llobregat.

IDIBELL is a member of the Campus of International Excellence of the University of Barcelona HUBc and is part of the CERCA institution of the Generalitat de Catalunya. In 2009 it became one of the first five Spanish research centers accredited as a health research institute by the Carlos III Health Institute. In addition, it is part of the “HR Excellence in Research” program of the European Union and is a member of EATRIS and REGIC. Since 2018, IDIBELL has been an Accredited Center of the AECC Scientific Foundation (FCAECC).

Scroll to Top