We hope this work proves to be a valuable tool for the coral research and restoration community, encouraging others to creatively apply machine learning to disease classification questions. 🧬

This approach enabled us to identify gene expression profiles that classify phenotypically similar diseases, which helps us not only to characterize the differences between two similar diseases but also to create a reliable model for identifying co-occurring infections along a reef. 🪸

The model was successful in classifying in situ corals with either white plague or SCTLD, with a classification rate of 98.95% (AUC) and an overall error rate of 0.0754. 🥳

Using coral gene expression data, we identified biomarkers through logistic regression and differential expression. We built a Partial Least Squares Discriminant Analysis (PLS-DA) model that reliably distinguishes between WP and SCTLD, which was tested and validated using in situ samples.

Would love to be added thanks! 🪸

Would love to be added!