 This paper introduces Immune Cell Abundance Identifier, Imusell AI, a novel method for accurately estimating the abundance of various immune cell types from gene expression data. The method was evaluated using both sequencing data and publicly available expression data, demonstrating its superior performance compared to existing methods. Additionally, the authors used Imusell AI to analyze immunotherapy data sets and found that the abundances of certain T-cell subsets were significantly higher in responders than in non-responders. Finally, they developed a model based on Imusell AI results which could accurately predict the response to immunotherapy. This article was authored by Yerou Miao, Chiung Zhang, Qian Lei, and others.