Data harmonization for large multi-site organizations

Harmonization in Cell Preparation: Preserving Data Comparability Across Sites, CROs, and Automation Platforms

By Mallory Griffin, Curiox Biosystems

The problem: comparable data starts upstream

Cell-based assays have benefited from major advances in instrumentation, reagents, and analysis software. Measurement technologies and data analysis are increasingly automated, standardized, and shared across organizations. Cell preparation, however, remains one of the least controlled parts of the workflow. 

Even when the same protocol is used, washing, media exchange, and handling steps are often executed differently across users and sites. These differences may be manageable within a single lab, but they become problematic when assays are shared across teams, transferred between locations, or repeated over time. As a result, variability attributed to biology or instrumentation often originates upstream, during sample preparation. Multiple interlaboratory studies have shown that pre-analytical handling contributes more variance than downstream analytical instrumentation in cell-based assays [1, 2]. 

As this reality has become harder to ignore, it has brought forward a concept that sits upstream of reproducibility: harmonization. The importance of harmonization for large, distributed scientific teams was recently highlighted in a July 2025 harmonization guide [3], which emphasized the need for shared execution definitions and supporting infrastructure to enable consistent implementation across groups and technologies. 

Harmonize Cell Washing with Software

Curiox’s C‑FREE Pluto Code packages centrifuge‑free washing logic into a software library for existing liquid handlers, so cell washing parameters can be reused across sites and applied consistently on different automation platforms.

C-FREE Pluto Code

This article focuses on harmonization at the level of cell preparation, where execution variability most often enters automated workflows.

What “harmonization” means scientifically

In cell preparation workflows, harmonization refers to the intentional alignment of execution parameters so that different operators or teams perform functionally equivalent workflows, enabling meaningful comparison of downstream data. These parameters include step order, timing, volumes, exchange behavior, and the rules that govern execution. 

Harmonization is often conflated with related concepts, but the distinctions matter: 

  • Reproducibility describes an observed outcome. 
  • Standardization typically refers to using the same tools or protocols. 
  • Harmonization addresses the process—whether workflows are designed to be equivalent in execution. 

Harmonization does not claim to improve biology. It is a design condition that constrains variability introduced by execution, making downstream comparisons more interpretable and transferable. This distinction aligns with how regulatory, clinical, and consortium frameworks treat reproducibility as an outcome, while harmonization is treated as a design requirement [4].

Why cell preparation resists harmonization

Cell preparation is difficult to harmonize because many of its most influential variables are implicit. Manual workflows encode critical parameters—aspiration behavior, mixing intensity, dwell times, judgments about completeness—through technique rather than specification. Standard operating procedures (SOPs) describe what to do, but rarely capture how actions are physically executed. As a result: 

  • Two operators following the same protocol may apply very different forces or timings 
  • The same workflow may be interpreted differently across teams 
  • Key execution variables remain invisible and non-transferable 

Studies evaluating aspiration speed, wash force, and resuspension intensity have demonstrated measurable impacts on cell recovery, stress markers, and population frequencies [5, 6, 7]. This is not a failure of rigor. Many preparatory techniques evolved as hands-on practices, optimized locally and passed on informally. That approach works—until workflows need to move. 

Importantly, this execution-dependent variability has been documented across cell-based assays. Studies in cytometry and assay development literature show that aspiration, washing, and resuspension steps can introduce measurable cell stress and variability even when downstream detection and analysis are well controlled.

Harmonization and reproducibility serve different roles

Reproducibility describes whether a result can be obtained again. Harmonization addresses whether a workflow is designed to produce comparable results when executed in different contexts. 

A workflow may be reproducible within a single lab without being harmonized across sites. Skilled operators and stable local conditions can yield consistent outcomes even when execution details remain implicit. When that same workflow is transferred—to another team, instrument, or site—results often diverge, not because the biology has changed, but because execution has. This is why reproducibility often fails during assay transfer—not because assays are invalid, but because execution assumptions are no longer shared [8]. 

Harmonization reduces this fragility by making preparation parameters explicit and system-defined. In this way, reproducibility becomes an outcome that can persist beyond a single lab rather than a local property dependent on technique.

What harmonized cell preparation looks like in practice

In biopharma, harmonization is not about enforcing identical protocols or tools. It is about making critical execution parameters explicit, controlled, and transferable, so assays behave comparably when run by different teams, at different sites, or on different automation platforms. 

  • Defined exchange behavior – How fluid is exchanged matters as much as how much. 

Instructions such as “remove supernatant” or “resuspend cells” conceal substantial variability. Aspiration rate, proximity to the cell layer, and fluid introduction influence cell loss, stress, and assay readouts—particularly in sensitive cell-based and flow cytometry workflows [9]. In harmonized preparation, exchange behavior is defined, not implied. 

  • Explicit timing and sequencingTiming is a parameter, not an operational convenience. 

In busy biopharma labs, timing between steps varies with workload and instrument availability. Small differences in dwell time or delays between washes can accumulate into meaningful variability. Harmonized workflows preserve step order and intervals so assays remain comparable across time and facilities. 

  • Fewer discretionary decision pointsReduce reliance on individual judgment during execution. 

Many preparation steps depend on operator judgment—when a wash is “complete,” how vigorously to resuspend, or when to proceed. Harmonized workflows encode execution rules directly into the process, reducing dependence on local technique and experience [10]. 

  • Parameter preservation across instruments – Equivalence matters more than uniformity.

Large organizations rarely standardize on a single automation platform. Large biopharma organizations routinely operate mixed automation fleets, making execution portability more practical than hardware uniformity [11, 12]. One site may use a Biomek system while another uses Opentrons or a different liquid handler. Harmonization does not require identical hardware; it requires that defining execution parameters (e.g. CQAs, Industry 5.0 reppraisal) are preserved across instruments, so the same preparation logic produces comparable results regardless of platform.

Why automation enables harmonization across sites and platforms

Automation enables harmonization by externalizing execution parameters from individual technique into the workflow itself. Exchange behavior, timing, sequencing, and progression rules become executable parameters rather than inferred actions. 

In biopharma settings, this is critical during assay handoff, tech transfer, or CRO execution, where execution details must be specified, transferred, and audited. Manual workflows struggle to support this level of consistency at scale. 

In practice, this can take different forms. For example, software-defined approaches such as Curiox’s C-FREE Pluto Code allow centrifuge-free washing to be deployed on existing liquid handlers, rather than requiring new standalone instruments. By packaging washing logic as a software library, execution parameters can be reused across sites and applied consistently on different platforms, simplifying automation layouts while supporting harmonized execution.

Why harmonization matters for assay transfer, CROs, and analytical groups

Once assays move beyond their point of origin, the consequences of non-harmonized preparation become unavoidable. 

In biopharma organizations, this often surfaces during assay transfer, tech transfer, or CRO execution. A workflow that performs reliably within one group may show subtle but persistent shifts when executed by another team, even when the same written protocol is followed. These differences are frequently discovered only after data has been generated—during comparability studies or downstream analysis—when they are costly to resolve. 

Analytical, QC, and assay development groups are particularly exposed to this risk. When preparation depends on local technique or tacit knowledge, downstream standardization in instrumentation or analysis cannot fully compensate. 

Harmonization addresses this problem upstream. By aligning execution parameters, harmonized preparation creates a stable foundation for comparing data across sites, instruments, and partners. It allows teams to distinguish variability introduced by biology from variability introduced by execution—enabling reproducibility to persist beyond a single lab and allowing data generated across teams and platforms to be interpreted together with confidence. 

A practical next step 

For organizations exploring how harmonized, centrifuge-free preparation can be implemented without increasing system complexity, software-defined approaches offer a practical path forward. 

Curiox’s C-FREE Pluto Code, enables centrifuge-free washing on existing liquid handlers through a software library installation. By defining washing behavior at the software level, teams can share execution parameters across sites without introducing additional hardware. 

Learn more about C-FREE Pluto Code →

Meet Curiox at SLAS

Talk with our team about removing the centrifuge from cell preparation to support harmonized automation workflows and more reliable assay transfer across sites and partners.

Visit our SLAS page to schedule a conversation

Meet Curiox at SLAS

Talk with our team about removing the centrifuge from cell preparation to support harmonized automation workflows and more reliable assay transfer across sites and partners.

Schedule a conversation at SLAS

References

  1. Czechowska K, Lannigan J, Aghaeepour N, Back JB, Begum J, Behbehani G, Bispo C, Bitoun D, Fernández AB, Boova ST, Brinkman RR, Ciccolella CO, Cotleur B, Davies D, Dela Cruz GV, Del Rio-Guerra R, Des Lauriers-Cox AM, Douagi I, Dumrese C, Bonilla Escobar DL, Estevam J, Ewald C, Fossum A, Gaudillière B, Green C, Groves C, Hall C, Haque Y, Hedrick MN, Hogg K, Hsieh EWY, Irish J, Lederer J, Leipold M, Lewis-Tuffin LJ, Litwin V, Lopez P, Nasdala I, Nedbal J, Ohlsson-Wilhelm BM, Price KM, Rahman AH, Rayanki R, Rieger AM, Robinson JP, Shapiro H, Sun YS, Tang VA, Tesfa L, Telford WG, Walker R, Welsh JA, Wheeler P, Tárnok A. Cyt-Geist: Current and Future Challenges in Cytometry: Reports of the CYTO 2019 Conference Workshops. Cytometry A. 2019 Dec;95(12):1236-1274. doi: 10.1002/cyto.a.23941. PMID: 31833655. 
  2. Maecker HT, Trotter J. Flow cytometry controls, instrument setup, and the determination of positivity. Cytometry A. 2006 Sep 1;69(9):1037-42. doi: 10.1002/cyto.a.20333. PMID: 16888771. 
  3. Ruiz, O. E., Wagenaar, J. B., Mehta, B., Ziogas, I., Swanson, L., Worley, K. C., Cruz-Almeida, Y., Johnson, A. J., Boline, J., Boccanfuso, J., RE-JOIN Consortium, Martone, M. E., & Haelterman, N. A. (2025). A guide to developing harmonized research workflows in a team science context. Experimental neurology, 392, 115333. https://doi.org/10.1016/j.expneurol.2025.115333 
  4. CLSI. Validation of Assays Performed by Flow Cytometry, 1st ed. CLSI guideline H62. Clinical and Laboratory Standards Institute; 2021. 
  5. Xie Y, Wang F, Puscheck EE, Rappolee DA. Pipetting causes shear stress and elevation of phosphorylated stress-activated protein kinase/jun kinase in preimplantation embryos. Mol Reprod Dev. 2007 Oct;74(10):1287-94. doi: 10.1002/mrd.20563. PMID: 17492777. 
  6. Sato M, Levesque MJ, Nerem RM. Micropipette aspiration of cultured bovine aortic endothelial cells exposed to shear stress. Arteriosclerosis. 1987 May-Jun;7(3):276-86. doi: 10.1161/01.atv.7.3.276. PMID: 3593075. 
  7. Chivukula VK, Krog BL, Nauseef JT, Henry MD, Vigmostad SC. Alterations in cancer cell mechanical properties after fluid shear stress exposure: a micropipette aspiration study. Cell Health Cytoskelet. 2015 Jan 9;7:25-35. doi: 10.2147/CHC.S71852. PMID: 25908902; PMCID: PMC4405123. 
  8. van Amsterdam P. Learning From Issues in Assay Cross-Validation / Method Transfers [conference presentation]. Presented at: 5th Bioanalysis Workshop; 2011 Apr 13-14; Montreal, Canada. Slide set uploaded to SlideShare by Peter van Amsterdam. 
  9. Oliveira B, Leite J, Daniels K, Gardner R. The Impact of Washes on Staining for Flow Cytometry [poster]. Presented at: CYTO Virtual 2021; 2021. Whitehead Institute for Biomedical Research & Memorial Sloan Kettering Cancer Center. Available from: https://wi.mit.edu/sites/default/files/2021-06/the_impact_of_washes.pdf 
  10. Begley CG, Ellis LM. Drug development: Raise standards for preclinical cancer research. Nature. 2012 Mar 28;483(7391):531-3. doi: 10.1038/483531a. PMID: 22460880. 
  11. Agilent Technologies Inc. Agilent automated workflows at SLAS2025: connecting with confidence [press release]. SANTA CLARA, Calif: Agilent Technologies Inc.; 24 Jan 2025. Available from: https://www.agilent.com/about/newsroom/presrel/2025/24jan-ca25003.html 
  12. Rathore AS, Sarin D. What should next-generation analytical platforms for biopharmaceutical production look like? Trends Biotechnol. 2024 Mar;42(3):282-292. doi: 10.1016/j.tibtech.2023.08.008. Epub 2023 Sep 27. PMID: 37775418. 
  13. Wolf Á, Zsoldos P, Széll K, Galambos P. Towards robotic laboratory automation plug & play: Reference architecture model for robot integration. SLAS Technol. 2024 Aug;29(4):100168. doi: 10.1016/j.slast.2024.100168. Epub 2024 Aug 2. PMID: 39098589.