Open qpisuo opened 1 week ago
Thanks for your interest! I’m updating the project to allow direct testing with the BiGG model so that the initial embeddings won’t be needed. I’ll post the update once it's complete.
Thank you for your reply! Is the BiGG model related to the biomedical field? I'm also interested in applying this algorithm to the drug discovery field. I will continue to follow your project. Thank you!
Hello! Thank you for sharing your work! While studying your project, I noticed an issue with the iAF1260b dataset. In the training data, there are some overlaps between positive and negative samples. I’m wondering if there might be an issue with the negative sample generation process. Thank you!