In our recent transcriptomic analysis of T. gondii, we used UMAP and HDBSCAN clustering to examine the organism's transcriptional landscape through its developmental stages. Initially, a low variance threshold led to many genes being labeled as noise and unclear clusters with low silhouette scores. To improve clarity, we set a stricter variance threshold of 2, reducing our dataset to 4,883 genes and enhancing the identification of distinct transcriptional signatures.

Investigating Regression model results and improving the model