Users are allowed to load data from multiple resources. We adopted data object strcuture desgin and pre-processing steps from Seurat V3. Users famliar with Seurat can use Seurat functions to load data and create objects. We also provided some functions for data loading.

Using “LinQ-View”, we are able to perform joint analysis for multi-modal datasets and compare the results with transcriptome or surface protein alone. We performed joint analysis on several public multi-modal datasets (see method section for details). Results showed joint analysis achieves better clustering quality and deeper resolutions.

To evaluate the compatibility of “LinQ-View” with conventional computational batch effects correction methods, we performed joint analysis using “LinQ-View” on two datasets of PBMCs from healthy donors (called PBMC 1K and PBMC 10K). We performed joint analysis on two independent runs: one with batch effects correction and one without as control. Here, we use Seurat V3 to remove the batch effects.