Loss Function in SpatialGlue

JinmiaoChenLab / SpatialGlue

SpatialGlue is a novel deep learning methods for spatial multi-omics data integration.

GNU Affero General Public License v3.0

35 stars 5 forks source link

Loss Function in SpatialGlue #4

Open liu8831790 opened 3 months ago

liu8831790 commented 3 months ago

Hi, I've been reading the SpatialGlue paper and noticed that the loss function is composed of three parts: Reconstruction loss, Correspondence loss, and Adversarial loss. While in the code, it appears that the loss function only includes the Reconstruction loss and Correspondence loss. I'm curious why there is this difference. Is there an implementation detail I'm missing or an alternative approach taken in the code?

Thank you for your help!

longyahui commented 3 months ago

Hi,

Thank you for your attention on our work.

The code is consistent with our SpatialGlue paper published in Nature Methods. Please refer to our latest article via https://www.nature.com/articles/s41592-024-02316-4. Thanks again.

Best regards,
Yahui