Amodio G, Giacomini G, Boeri L, et al. **T cell exhaustion and senescence signatures characterize and differentiate infertile men**
Amodio G, Giacomini G, Boeri L, et al. **T cell exhaustion and senescence signatures characterize and differentiate infertile men**
### Single-Cell RNA Sequencing Analysis ###
### Single-Cell RNA Sequencing Analysis ###
...
@@ -9,15 +9,20 @@ scRNAseq analysis was performed using a standard [Seurat](https://satijalab.org/
...
@@ -9,15 +9,20 @@ scRNAseq analysis was performed using a standard [Seurat](https://satijalab.org/
- Preprocessing and cell filtering
- Preprocessing and cell filtering
- Each sample was pre-processed and cells with mitochondrial RNA percentages higher than 10 and a number of features <1200or>6000, were filtered out. Samples were merged into a single Seurat dataset
- Each sample was pre-processed and cells with mitochondrial RNA percentages higher than 10 and a number of features <1200or>6000, were filtered out. Samples were merged into a single Seurat dataset
- Data was regressed out by passing UMI count, the percentage of mitochondrial genes, the difference between the cell cycle phases scores, as described in the Seurat [vignette](https://satijalab.org/seurat/articles/cell_cycle_vignette.html#alternate-workflow-1).
- Data was regressed out by passing UMI count, the percentage of mitochondrial genes, the difference between the cell cycle phases scores, as described in the Seurat [vignette](https://satijalab.org/seurat/articles/cell_cycle_vignette.html#alternate-workflow-1).
- Dimensionality reduction and Harmony batch removal:
- Dimensionality reduction and Harmony batch removal:
- A principal component analysis (PCA) with 100 principal components (PCs) was performed and a UMAP-representation as well as clusters were computed on the top 55 components (orig.ident as batch variable)
- A principal component analysis (PCA) with 100 principal components (PCs) was performed and a UMAP-representation as well as clusters were computed on the top 55 components (orig.ident as batch variable)
- Clustering
- Clustering:
- Markers identification
- K-nearest neighbor (KNN) graph was first constructed based on the Euclidean distance using the [FindNeighbors](https://satijalab.org/seurat/reference/findneighbors) function, with the KNN algorithm set to 20.
- Cluster annotation:
- The modularity optimization technique was applied using the Louvain algorithm through the [FindCluster](https://satijalab.org/seurat/reference/findclusters) function, with resolution parameters set to 1.2.
- Intra-cluster comparisons
- Markers identification:
- Marker genes for each cluster were identified using the [FindAllMarkers](https://satijalab.org/seurat/reference/findallmarkers) function with the logfc.threshold argument set to 0.25. Only genes expressed in at least 25% of cells in one of the compared clusters were considered (min.pct = 0.25). Genes with pvalues <1e10<sup>-5</sup> from the Wilcoxon Rank Sum test were considered as markers for a specific cluster.
- Cluster annotation
- Gene enrichment analysis (GSEA):
- Intra-cluster comparisons: Intra-cluster comparisons among the experimental conditions were conducted using the [FindMarkers](https://satijalab.org/seurat/reference/findmarkers) function, setting test.use = wilcox, a logFC threshold = 0, min.cells.group = 5 and return.thresh parameter equal to 1.
- GSEA function of [ClusterProfiler R package](https://bioconductor.org/packages/release/bioc/manuals/clusterProfiler/man/clusterProfiler.pdf) was applied, using the full marker gene list ranked by decreasing logFC and the hallmarks gene set. Gene sets were considered enriched if their adjusted pvalue was <0.1.
