← Back to article | PVAED: prior-guided variational autoencoders with diffusion denoising for interpretable single-cell representation learning