ScRNA-seq data processing

Droplet-based raw data were processed using Cell Ranger (Version 3.0.0)⁴⁶ against the GRCh37 human reference genome with default parameters. First, data from each batch was normalized separately using the NormalizeData function and scaled with the ScaleData function implemented in the Seurat pipeline²⁹. Then data from different batches were integrated using the canonical correlation analysis (CCA) method implemented in Seurat²⁹. For each subset of immune cells, NK cells and macrophage subtypes were identified based on the clustering analysis of the single-cell profiles from RPL patients and healthy control individuals shown in Fig. ​Fig.1.1. T cells were extracted, normalized, re-clustered, and analyzed separately. We retained cells with detected gene numbers between 500 and 3000 and less than 10% mitochondrial UMIs. Moreover, genes expressed in fewer than three cells were also excluded. Downstream data processing and analysis steps, including filtering, normalization, batch removal, dimension reduction, were performed using the Seurat package version 2.3.1. We clustered all the cells based on the integrated gene expression matrix using Seurat with a parameter Resolution=0.6 and generated 11 clusters. In addition, we combined the Droplet-based dataset from Vento-Tormo et al.²⁵ with our transcriptome data from healthy controls (1:20 subspaces) using CCA in the Seurat pipeline. To quantify the similarity of the two datasets, we applied matchSCore2⁴⁷ to calculate the Jaccard index of clusters using the top 100 ranking cell-type-specific marker genes. In addition to Seurat, we also used Harmony³⁰ to integrate different batches from healthy controls and RPL patients to verify our integration reliability. We used the same gene expression matrix as used in Seurat and performed RunHarmony function in Hamony with default parameters to perform data integration. We then used the same clustering algorithm as that used in Seurat to cluster the cells with resolution 0.3 to generate distinct cell type clusters.