Space Ranger provides pipelines for end to end analysis of Visium Spatial Gene Expression experiments:

• spaceranger mkfastq demultiplexes Visium-prepared raw base call (BCL) files generated by Illumina sequencers into FASTQ files. It is a wrapper around Illumina's bcl2fastq with additional features that are specific to 10x Genomics libraries and a simplified sample sheet format. Please note that you can also use BCL Convert or bcl2fastq to convert BCL files into FASTQ files.

• spaceranger count takes a microscope slide image and FASTQ files and performs alignment, tissue detection, fiducial detection, and barcode/UMI counting. The pipeline uses Visium Spatial Barcodes to generate feature-barcode matrices, determine clusters, and perform gene expression analysis. In addition to fresh-frozen tissues, spaceranger count is also used for single-library analysis of FFPE and Protein Expression.

• spaceranger aggr takes the output of multiple runs of spaceranger count from related samples and aggregates their output, normalizing those runs to the same sequencing depth, and then recomputing the feature-barcode matrices and the analysis on the combined data. The aggr pipeline combines data from multiple samples into an experiment-wide feature-barcode matrix and analysis. This pipeline is supported for both fresh-frozen and FFPE tissues, although the aggregated libraries must be of the same type.

• spaceranger targeted-compare compares a starting input library, referred to as the parent library, to its corresponding Targeted Gene Expression dataset. spaceranger targeted-compare can be used to assess targeting performance with greater accuracy than when only the targeted data are known. Quality control metrics are provided to verify the extent to which targeted genes were enriched and the parent sample data were recovered. This pipeline is only supported for fresh-frozen tissues.

• spaceranger targeted-depth summarizes a whole transcriptome analysis (WTA) dataset in the context of a hypothetical Targeted Gene Expression experiment. Given an existing WTA dataset and a target panel CSV file, spaceranger targeted-depth computes the fraction of reads mapped to targeted genes from the panel. This pipeline is only supported for fresh-frozen tissues.

Space Ranger takes a slide image as input to be used as an anatomical map on which gene expression measures are visualized. Space Ranger can take two styles of images: either a brightfield image stained with hematoxylin and eosin (H&E) with dark tissue on a light background, or a fluorescence image with bright signal on a dark background. While brightfield input comprises a single image, fluorescence input comprises one or more channels of information generated by separate excitations of the sample.

Space Ranger combines algorithms for processing Visium slide images with the robust gene expression analysis used in Space Ranger. The spaceranger count pipeline performs the following automated image processing tasks:

• Alignment of the slide's barcoded spot pattern to the input slide image. (Brightfield only)
• Discrimination between tissue and background in the slide image. (Brightfield only)
• Alignment of CytAssist image and microscope image of the same tissue section. (Optional, CytAssist Workflow only)
• Preparation of the full-resolution slide image for visualization in Loupe Browser.

These image processing tasks are described further in the Image Processing section. We encourage you to read Image Recommendations for imaging suggestions prior to data collection.

The automated alignment and tissue identification processes require images taken with proper tissue placement and image exposure and currently only works for brightfield images. Samples with fluorescence images or brightfield images where these conditions cannot be met can be processed by using manual fiducial alignment wizard in Loupe Browser to manually align the spot pattern and identify tissue in the slide image. When this manual workflow is chosen, the spaceranger count pipeline will still take the full resolution image or images as input in order to prepare data for visualization within Loupe Browser.

For the CytAssist workflow, the automatic image registration process will align the CytAssist image with the microscope image of the same tissue tissue captured on standard glass slide. A manual counterpart of this process is available in Loupe Browser as the CytAssist image alignment wizard where landmark-based point selection helps identify equivalent regions between the two image inputs.

Irrespective of the manual workflow performed, both the Loupe Manual Aligner and Space Ranger must be given the same set of images for a given sample.

Libraries supported by different versions of Space Ranger are summarized below.