Journal Article

Community assessment of methods to deconvolve cellular composition from bulk gene expression

Caruso, Francesca P., Coller, John, Guerrero-Gimenez, Martin E., Bertus, Thomas J., Ceccarelli, Michele, Saez-Rodriguez, Julio, Heiser, Laura M., Monaco, Gianni, Lamb, Andrew, Newman, Aaron M., Petitprez, Florent, Derrick, Daniel, Yu, Thomas, Shokoohi, Vida, Domanskyi, Sergii, Li, Haojun, Xiao, Xu, Yang, Wenxian, Maecker, Holden, Chung, Verena, Waterfall, Joshua J., Bletz, Julie, Wang, Shun, Duault, Caroline, Banerjee, Jineta, Piermarocchi, Carlo, Simon, Stockard, Lin, Yating, Lang, Benjamin J., White, Brian S., de Reyniès, Aurélien, Gentles, Andrew J., Guo, Xindi, Yu, Rongshan, Liliental, Joanna E., Catania, Carlos A., Valdeolivas, Alberto, Zheng, Frank, Guinney, Justin, Patel, Shailja

2024

Overview

We evaluate deconvolution methods, which infer levels of immune infiltration from bulk expression of tumor samples, through a community-wide DREAM Challenge. We assess six published and 22 community-contributed methods using in vitro and in silico transcriptional profiles of admixed cancer and healthy immune cells. Several published methods predict most cell types well, though they either were not trained to evaluate all functional CD8+ T cell states or do so with low accuracy. Several community-contributed methods address this gap, including a deep learning-based approach, whose strong performance establishes the applicability of this paradigm to deconvolution. Despite being developed largely using immune cells from healthy tissues, deconvolution methods predict levels of tumor-derived immune cells well. Our admixed and purified transcriptional profiles will be a valuable resource for developing deconvolution methods, including in response to common challenges we observe across methods, such as sensitive identification of functional CD4+ T cell states. Deconvolution methods infer levels of immune infiltration from bulk expression of tumour samples. Here, authors assess 6 published and 22 community-contributed methods via a DREAM Challenge using in vitro and in silico transcriptional profiles of admixed cancer and healthy immune cells.

• Publisher: Nature Publishing Group UK,Nature Publishing Group,Nature Portfolio
• Subject: 38/91 / 631/114/2397 / 631/1647/48 / 631/250 / 631/67/327 / 631/67/69 / Algorithms / Bulk sampling / Cancer / CD4 antigen / CD4-Positive T-Lymphocytes - metabolism / CD8 antigen / CD8-Positive T-Lymphocytes - metabolism / Cells / Computational Biology - methods / Deconvolution / Deep Learning / Gene expression / Gene Expression Profiling - methods / Gene Expression Regulation, Neoplastic / Humanities and Social Sciences / Humans / Immune system / In vitro methods and tests / Infiltration / Lymphocytes / Lymphocytes T / Lymphocytes, Tumor-Infiltrating - immunology / Metastases / multidisciplinary / Neoplasms - genetics / Neoplasms - immunology / Neoplasms - pathology / Science / Science (multidisciplinary) / Transcription / Transcriptome / Tumors