Single-Cell Loading Decision Framework: Confident Go/No-Go Before Chip Commitment

The Bottom Line Up Front: Single-cell chips cost hundreds of dollars. Loading poor-quality samples wastes that investment and produces compromised data. QC determines "whether a sample prep should be loaded at this point or should be cleaned up again". This decision framework transforms subjective quality judgments into objective load/clean/reject decisions - protecting expensive consumables while maximizing single-cell data quality.

Protecting Your Single-Cell Investment

Every single-cell experiment represents significant investment - consumables alone cost $500-1000+ per chip, plus sequencing costs downstream. Loading poor-quality samples wastes these investments and produces compromised data that may be unusable.

The decision framework enables confident go/no-go decisions based on objective metrics rather than hope.

TL;DR - Loading Decision Essentials

  • Pre-loading QC determines "whether a sample prep should be loaded at this point or should be cleaned up again"
  • LOAD: Debris <20%, Viability ≥80% - proceed with confidence
  • CLEAN: Elevated debris but good viability - cleanup will help
  • REJECT: High debris AND low viability - fundamental quality issue
  • Objective metrics replace guessing - protect chip investments with data

Complete Loading Decision Framework

Learn how to make confident load/clean/reject decisions that protect your single-cell investments and maximize data quality.

The Cost of Poor Loading Decisions

Single-cell experiments represent significant investments. Poor loading decisions waste those investments.

Costs of Loading Poor Samples

  • High debris loaded: Clogged chip → failed run → lost chip cost
  • High soup loaded: Contaminated data → computational corrections may fail
  • Low viability loaded: Dead cells waste capacity → fewer usable cells
  • Combined problems: Complete run failure → total investment loss

Costs of Unnecessary Cleanup

  • Cell loss: Cleanup removes some cells along with debris
  • Processing time: Delays experiments
  • Reagent costs: Additional consumables
  • Handling damage: May damage remaining cells
THE DECISION IMPERATIVE

The framework enables avoiding both problems - loading poor samples AND unnecessary cleanup. Objective metrics enable optimal decisions for each sample.

Key Metrics for Loading Decisions

Effective loading decisions require multiple metrics that together characterize sample quality.

Primary Decision Metrics

  • Debris percentage: Contamination, soup potential, clogging risk - target <20%
  • Viability percentage: Sample health, dissociation damage - target ≥80%
  • Viable cell count: Available material for loading - must be sufficient
  • Total concentration: Loading calculation input - platform-specific

Metric Interactions

  • High debris + low viability: Fundamental sample problem
  • High debris + high viability: Cleanup will help - debris removable
  • Low debris + low viability: Dissociation damage - cells may decline
  • Cell count: Must be sufficient after potential cleanup losses
CASSETTE SELECTION

Use S+ cassettes (3-27 μm) for most tissue-derived cells or M+ cassettes (4-34 μm) for larger cells.

The Three-Path Decision Framework

Every sample assessment leads to one of three decisions: Load, Clean, or Reject. QC metrics determine which path.

PATH 1: LOAD

  • Debris percentage: <20%
  • Viability percentage: ≥80%
  • Cell count: Sufficient for target
  • Action: Proceed to loading with confidence

PATH 2: CLEAN

  • Debris percentage: 20-40%
  • Viability percentage: ≥75%
  • Cell count: Sufficient for cleanup losses
  • Action: Perform cleanup → re-assess → load

PATH 3: REJECT/TROUBLESHOOT

  • Debris >40% AND Viability <70%
  • OR Cell count insufficient
  • OR Viability <60% regardless of debris
  • Action: Sample fundamentally compromised → troubleshoot source
PLATFORM ADJUSTMENTS

10x Genomics: Strict thresholds (debris <15%, viability ≥85%). FACS: Higher debris tolerance. Sort then load: Can compensate for higher initial debris.

Implementation Protocol

Pre-Loading QC Protocol

  1. Prepare Sample: Complete dissociation and any standard cleanup
  2. Collect Aliquot: Remove 50-100 μL for analysis
  3. Add Viability Dye: Stain for viability assessment
  4. Run Analysis: Use Moxi V or Moxi GO II with appropriate cassette
  5. Record Metrics: Total concentration, viable concentration, viability %, debris %
  6. Apply Framework: Compare metrics to thresholds → determine path
  7. Execute Decision:
    • Load: Calculate loading concentration → proceed
    • Clean: Perform cleanup → repeat QC → load
    • Reject: Document → troubleshoot source
  8. Document: Record QC metrics and decision for sample tracking
Edge Cases and Special Considerations

Real samples don't always fit neatly into categories. Edge cases require judgment informed by metrics.

Borderline Samples

  • Debris 18-22%, Viability 78-82%: Borderline on both - consider cleanup if sample allows, otherwise load with caution
  • High debris (30%), excellent viability (90%): Cleanup likely very effective - expect good outcome
  • Low debris (10%), borderline viability (75%): Dissociation damage - load quickly, viability may decline
  • Precious sample with marginal quality: Cannot repeat - document expectations, load with awareness

Tissue-Specific Adjustments

  • Brain tissue: Higher debris thresholds acceptable (inherently debris-prone)
  • Tumor tissue: Viability may be inherently lower (necrosis)
  • Frozen samples: Lower baseline viability expected
TIME SENSITIVITY

Viability may decline during extended processing. If cleanup takes significant time, recheck before loading. Prioritize speed with marginal samples.

Troubleshooting Loading Decisions

Problem: Most samples require cleanup or rejection
Solution: Upstream dissociation problems likely. Optimize dissociation protocol. Improve tissue handling. Consider tissue source quality. Review sample transport and storage. Use QC data to identify pattern.
Problem: Cleanup doesn't achieve load threshold
Solution: Cleanup method may be ineffective for debris type. Try different cleanup method. Optimize cleanup parameters. Accept cleaned sample if improvement shown. Troubleshoot source if persistent.
Problem: QC passed but single-cell run still failed
Solution: Minimize time between QC and loading. Consider debris composition (sticky debris may cause problems despite acceptable percentage). Review handling during loading. Track correlation between specific metrics and outcomes to refine thresholds.
Problem: Uncertainty about borderline samples
Solution: Consider sample value (precious vs. replaceable). Err toward cleanup if time allows. Document decision rationale. Track outcomes to refine personal thresholds over time.

Common Questions About Loading Decisions

How do I decide whether to load, clean, or reject a sample?
Quantitative QC enables objective decisions. Load if debris <20% and viability ≥80%. Clean if debris is elevated (20-40%) but cells healthy (viability ≥75%). Reject if both debris high (>40%) AND viability poor (<70%) - fundamental quality issue that cleanup won't fix.
What metrics matter most for single-cell loading?
Key metrics: debris percentage (affects soup and clogging), viable cell count (determines recovery), viability percentage (sample health), and total concentration (loading calculations). Together these predict run success and should all be evaluated.
Why can't I just load and filter computationally?
Computational filtering has limits. High debris creates ambient RNA "soup" that corrections struggle with. Dead cells waste chip capacity that cannot be recovered. Pre-loading QC determines "whether a sample prep should be loaded at this point or should be cleaned up again" - prevention beats correction.
What are the consequences of loading poor-quality samples?
Loading poor samples wastes expensive chips ($500-1000+), produces corrupted data requiring computational correction (which may fail), reduces usable cell recovery, and may require complete experiment repetition. QC investment prevents these costly outcomes.

Key Takeaway

Single-cell loading decisions should be data-driven, not hopeful. QC determines "whether a sample prep should be loaded at this point or should be cleaned up again". The three-path framework - Load, Clean, or Reject - transforms subjective judgments into objective decisions that protect chip investments, maximize data quality, and build institutional knowledge about sample preparation standards.
Back to all resources