This plot shows the predicted per-cell type TP10k utilization (a combination of all the genes' expression profiles for each cell type) based on the reference dataset you provided.

To ensure optimal results, check the utilization values for each cell type to be approximately below 1400 TP10k. High utilization for a cell type can lead to optical crowding which results in reduced detection sensitivity (the fraction of transcripts detected per cell) and limits accurate quantification of lowly expressed genes in affected cells.

This plot shows the predicted per-gene TP10k utilization alongside the corresponding cell types associated with each gene. You can change how many genes are shown by using the slider, up to the top 25 most highly expressed genes per cell type.

To ensure optimal results, check that the utilization values for each gene to be below approximately 120 TP10k. High utilization for a gene can lead to optical crowding which results in reduced detection sensitivity and limits accurate quantification of lowly expressed genes in affected cells.

We also recommend that you do not include genes which are moderately expressed in a large number of cell types, as doing so can limit the optical budget available in those cell types without providing much information.

This section shows if you have genes on your panel that you may wish to reconsider on your Xenium panel. We currently test for:

1. Extremely high expressors in TCGA data for tissues/conditions relevant to your selection. This is often a good indicator that a gene will cause optical crowding issues
2. Genes which have caused degraded assay performance in internal panel testings by 10x Genomics

This plot shows the number of probesets for each gene on your panel. The effectiveness of gene detection and subsequent sensitivity is correlated with the number of probesets used.

To ensure your genes have robust detection, make sure they have at least three probesets.

For details on how probesets are assigned to genes, please view our support site. In some instances, the number of probesets will be less than the default of 8 because we either could not design that many probesets for a gene, or because one or more of the probesets were removed as they were predicted to interact with other probes on your panel. We show this, and any recommended changes to the requested number of probesets, using colored outlines in the following plot.

Recommended optimization: These are genes with probesets which we recommend adjusting in order to achieve an optimal optical budget in your final panel. This will generally reduce the sensitivity of the individual gene, but ensure other co-expressed genes retain high sensitivity by limiting optical crowding.

At probeset limit: These are genes which are using all available probesets designed by 10x Genomics.

This plot shows hierarchical clustering of the single cell data you selected, subset to genes on your panel that are expressed. A higher z-score (calculated across cell types) indicates that the gene is expressed at a higher level in that cell type compared to other cell types in the dataset.

To ensure optimal results, you should make sure that:

1. Each cell type forms a distinct cluster (a different pattern per row), indicating that you will likely be able to identify the relevant cell types in your Xenium data
2. Each cell type has an appropriate number of marker genes (the number may vary depending on how deeply you intend to characterize the cell type)
3. A gene is not expressed uniformly across all cell types (i.e. not the same color in all rows)
4. Your panel genes are present in each single cell reference dataset you provided