ESTIMATE algorithm is not appropriate for inferring tumor purity and stromal and immune cell admixture in hematopoietic or stromal tumors

Excerpt

With great interest, we have read the recent article entitled “Identification of prognostic genes in the acute myeloid leukemia immune microenvironment based on TCGA data analysis” [1] (by Yan et al. Cancer Immunol Immunother (2019) 68: 1971), which tried to identify key microenvironment-related genes correlated with prognosis of acute myeloid leukemia by analyzing publicly available gene expression data. The authors calculated immune/stromal scores for all acute myeloid leukemia (AML) patients based on Estimation of STromal and Immune cells in Malignant Tumors using Expression data (ESTIMATE) algorithm, and they found that their calculated immune and stromal scores were associated with the subtype classification of AML patients. Next, the authors identified differentially expressed genes according to immune scores and stromal scores, and further found that key genes seem to have prognostic value for AML patients. Although we appreciate efforts of the authors for data analysis and collection, some evident mistakes were found out in this article. …