Tips:
In this module, we designed a feature that allows users to upload custom gene sets or specific signaling pathway genes for analysis within our curated datasets. After selecting the cancer type, dataset, tissue type, and cell type, and uploading the gene sets, the system generates a boxplot displaying gene set activity both before and after ICB treatment based on the available data. Each point in the boxplot represents an individual sample. For both pre and post group, samples are further categorized as responders or non-responders, allowing for comparison of gene set activity between these groups. It should be noted that sample sizes for each dataset vary considerably, and some datasets may have too few samples to perform statistical tests across the described groups. To address this, we implemented a pseudo-bulk strategy where 500 cells are randomly sampled from each dataset 20 times. For each iteration, we calculate the mean activity value for the respective biopsy sample. This method effectively increases the sample size based on the original dataset, enabling the identification of significantly differentially expressed gene sets between groups.