Bisque, a new computational tool, accurately estimates cell type proportions from bulk gene expression data by leveraging single-cell information and correcting for biases between sequencing methods. Tested on both adipose and brain tissue samples, Bisque outperformed existing tools and replicated known links between cell proportions and clinical traits like BMI and insulin resistance.

None identified

The paper presents a novel method, Bisque, for estimating cell type proportions in bulk expression data by integrating single-cell information. The method addresses a significant challenge in transcriptomics by accounting for technical biases between different sequencing technologies. The evaluation on adipose tissue and cortex datasets shows improved performance compared to existing methods and replicates previously reported associations between cell type proportions and phenotypes. The efficiency of the method and the alternative marker-based approach are also valuable contributions. While certain limitations exist regarding the generalizability and assumptions about cell population representation, the overall methodology and findings are strong, warranting a rating of 4.