Benchmarking cell-type deconvolution in cross-platform transcriptomic data.

SpatialDecon and cell2location achieve the most reliable cross-platform cell-type deconvolution, maintaining accuracy despite technological biases in transcriptomic data.

• Why it matters: Accurate tissue composition analysis is crucial for understanding biological heterogeneity, but platform-specific biases hinder consistent deconvolution across different technologies.
• What they did: The study systematically benchmarked deconvolution methods using real and simulated datasets that mimic diverse technological features, assessing performance across platforms.
• The result: Findings guide researchers in choosing appropriate computational tools based on experimental design, enabling more robust and consistent tissue deconvolution across transcriptomic platforms.