COOBoostR: An Extreme Gradient Boosting-Based Tool for Robust Tissue or Cell-of-Origin Prediction of Tumors

Abstract

We present here COOBoostR, a computational method designed for the putative prediction of the tissue- or cell-of-origin of various cancer types. COOBoostR leverages regional somatic mutation density information and chromatin mark features to be applied to an extreme gradient boosting-based machine-learning algorithm. COOBoostR ranks chromatin marks from various tissue and cell types, which best explain the somatic mutation density landscape of any sample of interest. A specific tissue or cell type matching the chromatin mark feature with highest explanatory power is designated as a potential tissue- or cell-of-origin. Through integrating either ChIP-seq based chromatin data, along with regional somatic mutation density data derived from normal cells/tissue, precancerous lesions, and cancer types, we show that COOBoostR outperforms existing random forest-based methods in prediction speed, with comparable or better tissue or cell-of-origin prediction performance (prediction accuracy—normal cells/tissue: 76.99%, precancerous lesions: 95.65%, cancer cells: 89.39%). In addition, our results suggest a dynamic somatic mutation accumulation at the normal tissue or cell stage which could be intertwined with the changes in open chromatin marks and enhancer sites. These results further represent chromatin marks shaping the somatic mutation landscape at the early stage of mutation accumulation, possibly even before the initiation of precancerous lesions or neoplasia.

Colon cancer arises from differentiated cell lineages in the context of inflammation

“…As shown in the Fig. 6a (COOBoostR 7 , see Methods), while the majority of the sporadic colon cancers appear to originate from stem cells (52%), among the IBD-related cases goblet (40%) and Best4 (28%) cells represent the prevalent cells of origin. Moreover, a substantial proportion of the sporadic cases is also predicted to originate from goblet (22%) and Best4 cells (17%).…”

“…Collectively, our results reveal an alternative bottom-up route to intestinal tumorigenesis originating from Paneth cells in the murine small intestine and, allegedly, from secretory lineages in the human colon triggered by inflammatory cues as in IBD or through western-style dietary factors. To validate the relevance of our mouse study in patient-derived colon cancer, we took advantage of novel computational methods (47)(48)(49) developed to predict the cell of origin of tumors by matching the mutational density along the cancer genome with the profiles of epigenetic modifications characteristic of normal cell types (Suppl. Fig.…”

“…d . Barplot showing the predicted cell-of-origin in a cohort of IBD-CRC (N = 25) and sCRC (N = 257)( 35 ) based on the COOBoostR computational approach( 49 ) (see Methods). P value depicts result of the Fisher exact test.…”

“…10c-d). Hence, we compared by COOBoostR( 49 ) (see Methods) the individual mutational landscapes with the epithelial cell types of the colon to compute the putative cells of origin. As shown in Fig.…”

“…The latter can be explained by the low-grade, subclinical inflammation induced by western-style dietary habits, the major colon cancer risk factor, likely to underlie the de-differentiation of secretory lineages of the colon and their acquisition of stem-like and tumor-initiating features. To validate these observations, we applied novel computational methods [5][6][7] to predict the cell of origin of tumors by matching their genomewide mutation spectrum with the epigenetic landscapes of colonic cell lineages. While the majority of sporadic colon cancers are predicted to originate from stem cells, in patients with an IBD history, the predominant cell of origin switches to secretory lineages such as goblet cells.Taken together, our results show that, in the context of inflammation, intestinal cancer arises through the dedifferentiation of committed secretory lineages and the activation of the revival stem cell state.…”