Treeomics: A Mathematical Tool for Predicting Metastases
Researchers recently created a tool, Treeomics, to reconstruct the evolutionary pathway of metastases and to chart subclones to their anatomical locations.
Precision medicine to address the treatment of metastatic disease faces unique challenges. Some areas requiring elucidation are how tumors develop the ability to metastasize, how many primary tumor cells can give rise to metastases, and the spatial, evolutionary, and chronological factors affecting seeding of metastases at anatomical sites distinct from the primary tumor.
Researchers recently created a tool, Treeomics, to reconstruct the evolutionary pathway of metastases and to chart subclones to their anatomical locations. The developers of Treeomics focused on metastases of pancreatic, prostate, and ovarian cancers.1 This new tool recreates the seeding patterns of metastases and infers the ancestral subclones that seeded metastases at different anatomic locations. To do this, Treeomics uses several samples from various anatomic sites and assumes primarily monoclonal samples.
To develop Treeomics, researchers obtained sequencing data from genomic DNA from 26 tumor samples in 3 patients. Six of the samples were from primary tumors, and 20 samples were from metastases. The researchers also used genomic DNA from the normal tissue of the patients to enable identification of somatic variants.
Sequencing revealed 127,597 putative variants in both coding and non-coding regions, averaging 4908 variants per sample. Average sequencing coverage was at a depth of 69 times, and 97.5% of samples were sequenced at a depth greater than 10 times.
Filtering these results for quality generated 2105 putative variants. A targeted sequencing approach on these 2105 variants aligned, processed, and validated 381 of these mutations at an average sequencing depth of 731 times.