Question: Which metrics to use for troubleshooting issues in a Cell Surface Protein experiment?
Answer: The Library metrics for antibody application provide information that help to troubleshoot certain issues in a Cell Surface Protein experiment. In this article we provide general guidance on key metrics and suggested limits that can help in the troubleshooting process.
|
Key Antibody Metrics |
Description and Suggested Limits |
Possible Reasons for Failure |
| Antibody Reads in Cells (cellranger count) OR Fraction of Reads in cells (cellranger multi) |
The fraction of valid-barcode, valid-UMI, recognized antibody-barcode reads with cell-associated barcodes. Ideal > 50% |
Possible causes for the low value of antibody reads in cells:
|
|
Fraction of Antibody Reads |
Fraction of Antibody library reads that contain a recognized antibody barcode when aligned to the feature reference. In other words this is the mapping rate of antibody reads to the feature reference. Ideal > 50% |
|
|
Fraction of Antibody Reads Usable |
The reads should have a cell-associated barcode, a valid UMI and align to the feature reference. >20% or 1000 usable reads per cell is an acceptable limit |
This metric can be impacted by low antibody reads in cells and low fraction antibody reads. |
|
Fraction of Unrecognized Antibody Reads* |
Fraction of antibody reads with unrecognized barcode when aligned to the feature reference. Ideal <50% |
A high value could usually indicate that an inaccurate feature reference (or) wrong FASTQ reads were used. If the input files are correct, the warning could be caused by potential workflow issues:
Please contact support@10xgenomics.com for assistance. |
|
Fraction of Reads in barcodes with high UMI counts |
Fraction of the Antibody library reads that were lost after removing barcodes with unusually high UMI counts (possibly aggregates). Ideal < 5% |
Potential protein aggregates which might be caused by poor sample quality, poor washes, suboptimal staining and antibody titration processes. Pre-spinning the antibodies is critical for removing protein aggregates as well. |
*Available on Cell Ranger v7.1 and earlier. In Cell Ranger v7.2 and later, this can be inferred using the Fraction of Antibody reads metric. E.g., if Fraction of Antibody Reads metric is 50%, the corresponding Fraction of Unrecognized Reads metric will be 50%.
Typically these issues are shown as warnings or alerts in the web_summary.html file. Below is an example with the warnings of “low fraction of antibody reads usable” and “high fraction of unrecognized antibodies”. For this example, the warnings are caused by an incorrect feature reference CSV file that contained antibodies that do not match the experiment.
To achieve a successful cell surface protein experiment, we recommend following the Demonstrated Protocol: Cell Surface Protein Labeling for Single Cell RNA Sequencing Protocol for guidance on how to perform labeling and the following article for additional guidance: How can I optimize my TotalSeq™ antibody labeling protocol?
Products: Single Cell Gene Expression, Single Cell Immune Profiling