DECIPHER integrates disentangled representation learning and prototype-based cell-type deconvolution across molecular modalities
Wenpu Lai, Chenyang Li, Qin Deng, Yue Zhu, Chenyu Liu, Zhenhua Li, Oscar Junhong Luo
DECIPHER improves cell-type deconvolution across molecular modalities by combining disentangled representation learning with prototype-based methods, showing robust performance in simulated and real datasets.
bioRxiv · Sep 30, 2026 · PreprintBioinformatics