The Pre-Loading QC Checkpoint That Saves Expensive Runs
The High-Stakes Moment Before Loading
Single-cell genomics workflows have a critical decision point: do you load this sample onto an expensive chip, or does it need more cleanup? This go/no-go decision is typically made with incomplete information - cell count and viability, but rarely debris assessment. And debris creates the ambient RNA "soup" that complicates everything downstream.
The cost of loading a debris-contaminated sample extends far beyond the chip itself. Bioinformatics teams struggle with ambient RNA correction. Data quality suffers. Biological conclusions become uncertain. One pre-loading debris measurement can prevent all of this - transforming a blind decision into an informed one.
TL;DR - Pre-Loading QC Essentials
- High debris correlates with high ambient RNA - the "soup" that contaminates single-cell data
- Pre-loading debris assessment enables load vs. cleanup decisions
- Moxi debris-to-cell index functions as a soup detector
- Sample prep that passes debris QC produces cleaner sequencing data
- 60 seconds of QC can save thousands in wasted runs and rework
Implementing Pre-Loading QC
Learn how to establish debris checkpoints that protect your single-cell investments and improve downstream data quality.
Understanding the Ambient RNA Problem
Single-cell RNA sequencing assumes each droplet contains RNA from one cell. But cell preparations contain more than cells - debris, dead cell contents, and free-floating RNA create an ambient background that ends up in every droplet.
What Creates the "Soup"
- Dead cell lysis: Dying cells release their RNA contents
- Tissue processing: Dissociation methods rupture cells
- Debris fragmentation: Membrane fragments carry RNA
- Time delays: Sample degradation releases more material
High ambient RNA creates a background signal that must be computationally removed. This correction is never perfect - the more soup in your sample, the more uncertainty in your final data. Bioinformaticians spend hours on ambient RNA correction that could have been prevented by sample QC.
Downstream Consequences
Ambient RNA contamination doesn't just create noise - it creates systematic bias. Highly expressed genes contribute more to the soup, creating spurious correlations. Cell type assignments become uncertain. Differential expression analysis loses power. All because the soup was loaded onto the chip.
While computational methods can partially correct for ambient RNA, prevention is always better. A pre-loading debris check takes 60 seconds and can save hours of bioinformatics correction - or prevent uncorrectable data quality issues entirely.
Debris as a Soup Indicator
Debris and ambient RNA share a common source: cellular disruption. When cells break down, they release both debris (membrane fragments, aggregates) and free RNA. Measuring debris provides a practical indicator of soup contamination.
The Moxi "Soup Detector" Concept
"We talk about Moxi. There's a line in here that talks about [Moxi] is a soup detector". The debris-to-cell ratio measured by impedance correlates with ambient RNA levels - high debris typically means high soup contamination.
Moxi measures particle size distribution using the Coulter principle. Intact cells produce characteristic impedance signatures in a specific size range. Debris produces smaller signatures. The ratio of debris to cells provides a quantitative soup indicator without requiring RNA measurement.
Debris-to-Cell Index
| Cell Percentage | Soup Risk | Recommendation |
|---|---|---|
| >85% | Low | Proceed to loading |
| 70-85% | Moderate | Consider cleanup |
| <70% | High | Cleanup required |
These thresholds are starting points - calibrate for your specific workflow by correlating debris percentage with downstream data quality. Some applications tolerate more soup than others; establish thresholds based on your quality requirements.
Setting Load vs. Cleanup Thresholds
The goal of pre-loading QC is a clear go/no-go decision. Establishing thresholds transforms subjective assessment ("this sample looks dirty") into objective decision-making ("this sample is at 72% cells, below our 85% threshold").
Factors Affecting Threshold Selection
- Sample type: Primary cells tolerate more debris than cell lines
- Platform requirements: Different single-cell systems have different sensitivities
- Analysis goals: Rare cell detection requires cleaner samples
- Bioinformatics capacity: Robust pipelines can handle more soup
Run samples with varying debris levels through your complete workflow. Track final data quality metrics (cells recovered, ambient RNA fraction, cluster resolution). Identify the debris level above which quality degrades. Set your threshold conservatively below that point.
Documentation for Threshold Validation
- Sample set: Test 10+ samples spanning debris range
- Pre-load QC: Record debris percentage for each
- Run data: Capture cell recovery and quality metrics
- Correlation analysis: Plot debris vs. data quality
- Threshold determination: Identify quality failure point
- SOP integration: Document threshold and rationale
For typical single-cell preps, use S+ cassettes (3-27 um) on Moxi V or GO II - appropriate for most cell types loaded onto single-cell platforms. Larger cells may require M+ (4-34 um).
Making the Cleanup Decision
When samples fail debris QC, the decision isn't just "yes or no" - it's "cleanup and proceed vs. abort." This decision tree should be defined in advance, not made under pressure.
Cleanup Options
- Dead cell removal: Magnetic bead-based removal of dead cells and debris
- FACS sorting: High-purity sorting (if equipment available)
- Density gradient: Debris separation by density
- Wash/resuspension: Simple dilution may reduce debris concentration
"They loved that you could actually quantify the debris percentage because that was their whole platform was debris removal and they didn't have a good way to really quantify the debris. But in the Moxi you can see the pre and post cleanup". Run Moxi before AND after cleanup to verify improvement.
When Not to Cleanup
Cleanup procedures have costs: time, cell loss, and potential selection bias. Sometimes the better decision is to accept suboptimal samples or repeat the preparation entirely. Consider:
- Cell recovery from cleanup vs. expected benefit
- Time constraints and sample viability
- Whether debris is intrinsic to the sample type
- Downstream tolerance for remaining contamination
Multiple cleanup steps can alter cell populations through selective loss. If a single cleanup doesn't bring debris below threshold, consider whether the preparation method itself needs optimization rather than adding more cleanup rounds.
Integrating Pre-Loading QC into Workflow
For pre-loading QC to be effective, it must be mandatory and documented - not optional best practice. Integration into your single-cell workflow ensures consistent application across all samples and operators.
Workflow Position
Pre-loading debris QC should occur:
- After final preparation: Sample is ready to load
- Before chip commitment: Haven't used expensive consumables
- With time for cleanup: Can still act on results
- At consistent timing: Sample age affects debris
Debris accumulates over time as cells die and degrade. QC timing should be standardized - measure at a consistent point before loading, typically 15-30 minutes before chip loading begins.
SOP Language Example
"Prior to single-cell chip loading, assess sample quality using Moxi debris quantification. Record debris percentage in sample log. Samples with cell percentage ≥85% may proceed to loading. Samples with cell percentage <85% must undergo dead cell removal protocol and re-assessment before loading. Document all QC results regardless of pass/fail status."
A 60-second debris measurement costs essentially nothing. A failed single-cell run costs thousands in reagents alone, plus analysis time and project delays. Pre-loading QC has essentially infinite ROI when it prevents even one failed run.