Accurately resolving the composition of tumor-infiltrating leukocytes is pivotal for advancing cancer immunotherapy strategies. Despite the success of some clinical trials, applying these strategies remains limited due to the challenges in deciphering the immune microenvironment. In this study, we developed a streamlined, two-step workflow to address the complexity of bioinformatics processes involved in analyzing immune cell composition from transcriptomics data. Our dockerized toolkit, DOCexpress_fastqc, integrates the hisat2-stringtie pipeline with customized scripts within Galaxy/Docker environments, facilitating RNA sequencing (RNA-seq) gene expression profiling. The output from DOCexpress_fastqc is seamlessly formatted with mySORT, a web application that employs a deconvolution algorithm to determine the immune content across 21 cell subclasses. We validated mySORT using synthetic pseudo-bulk data derived from single-cell RNA sequencing (scRNA-seq) datasets. Our predictions exhibit strong concordance with the ground-truth immune cell composition, achieving Pearson's correlation coefficients of 0.871 in melanoma patients and 0.775 in head and neck cancer patients. Additionally, mySORT outperforms existing methods like CIBERSORT in accuracy and provides a wide range of data visualization features, such as hierarchical clustering and cell complexity plots. The toolkit and web application are freely available for the research community, providing enhanced resolution for conventional bulk RNA sequencing data and facilitating the analysis of immune microenvironment responses in immunotherapy. The mySORT demo website and Docker image are free at https://mysort.iis.sinica.edu.tw and https://hub.docker.com/r/lsbnb/mysort_2022 .
Keywords: Alpha diversity; Beta diversity; Cancer; Deconvolution; Immune microenvironment; Immunotherapy; Precision medicine.
© 2025. The Author(s).