The I3H Pipeline

Every Step Documented, Every Metric Tracked

Click any stage to expand its QC parameters, processing details, and quality benchmarks.

01
๐Ÿฉธ
T + 0 min ยท Collection

Peripheral Blood Draw & PBMC Isolation

Whole blood collected via standard phlebotomy, PBMCs isolated through density gradient centrifugation within the processing window.

Processing Windowโ‰ค 4 hrsDraw to isolation start
Typical Yield1โ€“3 ร— 10โถPBMCs per mL whole blood
Collection TubesCPT / EDTAStandardized across cohorts

Standardized collection protocols ensure consistency across all patient cohorts. Samples are processed at I3H's dedicated facility with tracked chain-of-custody from draw to cryopreservation. Temperature controlled at room temp during transit โ€“ no cold-chain artifacts that would alter cell surface marker expression.

  • โœ“ Ficoll-Paque density gradient centrifugation
  • โœ“ Automated cell counting via hemocytometer / Countess
  • โœ“ Red blood cell lysis QC (visual + scatter verification)
  • โœ“ Sample metadata captured at point of collection
02
๐Ÿ”ฌ
T + 4 hrs ยท Viability QC

Sample Viability & Cell Quality Assessment

Rigorous viability gating ensures only high-quality, representative samples enter the profiling pipeline.

Viability Thresholdโ‰ฅ 85%Minimum for pipeline entry
Typical Viability92โ€“96%Fresh PBMC average
Post-Thaw Viability88โ€“93%Cryopreserved samples

Viability is the single most critical QC checkpoint. Dead cells bind antibodies non-specifically, corrupting downstream data. I3H maintains viability well above the 85% minimum threshold through rapid processing, controlled thawing, and cisplatin-based dead cell exclusion during CyTOF staining.

  • โœ“ Cisplatin (195Pt) live/dead discrimination on CyTOF
  • โœ“ Trypan blue exclusion for pre-stain counting
  • โœ“ Controlled thaw protocol for cryopreserved samples
  • โœ“ Viability gate applied before all downstream analysis
  • โœ“ Samples below threshold flagged and excluded with documentation
03
๐Ÿงช
T + 5 hrs ยท MDIPA Staining

Maxpar Direct Immune Profiling Assay (MDIPA)

Standardized lyophilized antibody cocktail โ€“ 30+ metal-conjugated markers measuring surface and intracellular proteins at single-cell resolution.

Panel Size30+Metal-conjugated markers
Populations Resolved50Immune cell subsets
VendorStandard BioToolsSole source โ€“ lyophilized cocktail

The MDIPA kit is the backbone of I3H's standardized profiling. As a pre-validated, lyophilized antibody cocktail, it eliminates lot-to-lot variability โ€“ critical for comparing results across thousands of patient visits spanning years.

  • โœ“ No spectral overlap โ€“ mass cytometry uses metal isotope tags
  • โœ“ Batch-matched reagents across all cohorts
  • โœ“ Lineage, activation, and functional markers in single tube
  • โœ“ Intercalator (iridium) DNA stain for singlet discrimination
  • โœ“ EQ Four Element Calibration Beads for signal normalization
CD3CD4CD8CD19CD56CD14CD16CD45RACCR7CD127CD25HLA-DRCD38CD27IgD+ more
04
โšก
T + 8 hrs ยท CyTOF Acquisition

Mass Cytometry Data Acquisition

Cells are nebulized, ionized, and analyzed by time-of-flight mass spectrometry โ€“ 30+ parameters per cell at thousands of events per second.

Events Targetโ‰ฅ 200KLive singlet events per sample
Acquisition Rate300โ€“500Events / second
Output FormatFCS 3.1Flow Cytometry Standard

CyTOF acquisition is destructive โ€“ each cell is vaporized into heavy-metal ions and analyzed by mass spectrometry. This eliminates spectral overlap entirely. EQ beads run continuously for real-time calibration.

  • โœ“ EQ bead normalization across acquisition runs
  • โœ“ Gaussian discrimination parameters for doublet removal
  • โœ“ Instrument tuning verified before each batch
  • โœ“ Signal stability monitored throughout acquisition
05
๐Ÿงน
T + 12 hrs ยท Data Processing

Bead Normalization, Gating & QC

Raw FCS files undergo bead normalization, debris removal, and standardized gating before entering the analysis pipeline.

FCS Files Processed4,000+Annotated and QC'd
Gating StrategyAutomatedValidated against manual expert
CorrelationHighAuto vs. immunologist gating

I3H has developed a computational analysis approach that aligns with how immunologists think about the data. Validated against manual expert gating with strong correlation โ€“ scalable without sacrificing immunological rigor.

  • โœ“ EQ bead removal and signal normalization
  • โœ“ Live/dead gating (cisplatin-negative)
  • โœ“ Singlet gating (DNA intercalator)
  • โœ“ CD45+ immune cell gating
  • โœ“ Automated population identification across 50 subsets
  • โœ“ Batch effect monitoring across acquisition runs
06
๐Ÿ—„๏ธ
T + 24 hrs ยท Data Integration

Pennsieve Platform โ€“ Structured Data Management

Processed FCS files, annotations, and patient metadata organized in Pennsieve for discoverable, FAIR-compliant data management.

Data Objects4,000+Annotated FCS files
Metadata LayersClinical + LabLinked to patient records
Access ControlRole-BasedHIPAA-aligned permissions

Pennsieve provides the data management backbone โ€“ each FCS file is linked to its clinical metadata, processing provenance, and QC annotations, creating a queryable, cohort-spanning dataset.

  • โœ“ FAIR data principles (Findable, Accessible, Interoperable, Reusable)
  • โœ“ Provenance tracking from blood draw through analysis
  • โœ“ Cross-cohort data discovery and comparison
  • โœ“ Integration with Penn EMR systems
  • โœ“ API access for computational workflows
07
๐Ÿ“Š
T + 48 hrs ยท Analytics

Immune Fingerprinting & Informatics

Comprehensive immune fingerprints โ€“ quantifying 50 populations to establish baselines, detect perturbations, and track longitudinal trajectories.

Analysis OutputImmune Fingerprint50-population quantification
Cohort ComparisonCross-StudyStandardized by MDIPA protocol
IntegrationMulti-ModalCyTOF + EMR + Clinical

The final layer transforms population frequencies into actionable immune health profiles. Standardized upstream means fingerprints are directly comparable across patients, time points, and disease contexts.

  • โœ“ Population frequency quantification across 50 immune subsets
  • โœ“ Activation state profiling (HLA-DR, CD38, Ki-67)
  • โœ“ Longitudinal trajectory tracking per patient
  • โœ“ Cohort-level distribution analysis
  • โœ“ Integration with clinical outcomes data
  • โœ“ Dashboard visualization for investigators
Why This Matters

One Pipeline. One Panel. Thousands of Comparable Profiles.

4,000+Visits profiled with identical protocol
50Immune populations per fingerprint
< 48hBlood draw to analyzed fingerprint
0%Spectral overlap (mass cytometry)
โ‰ฅ 85%Viability threshold for pipeline entry
FAIRData managed on Pennsieve platform