Assessment of Automated Flow Cytometry Data Analysis Tools within Cell and Gene Therapy Manufacturing

Abstract

Flow cytometry is widely used within the manufacturing of cell and gene therapies to measure and characterise cells. Conventional manual data analysis relies heavily on operator judgement, presenting a major source of variation that can adversely impact the quality and predictive potential of therapies given to patients. Computational tools have the capacity to minimise operator variation and bias in flow cytometry data analysis; however, in many cases, confidence in these technologies has yet to be fully established mirrored by aspects of regulatory concern. Here, we employed synthetic flow cytometry datasets containing controlled population characteristics of separation, and normal/skew distributions to investigate the accuracy and reproducibility of six cell population identification tools, each of which implement different unsupervised clustering algorithms: Flock2, flowMeans, FlowSOM, PhenoGraph, SPADE3 and SWIFT (density-based, k-means, self-organising map, k-nearest neighbour, deterministic k-means, and model-based clustering, respectively). We found that outputs from software analysing the same reference synthetic dataset vary considerably and accuracy deteriorates as the cluster separation index falls below zero. Consequently, as clusters begin to merge, the flowMeans and Flock2 software platforms struggle to identify target clusters more than other platforms. Moreover, the presence of skewed cell populations resulted in poor performance from SWIFT, though FlowSOM, PhenoGraph and SPADE3 were relatively unaffected in comparison. These findings illustrate how novel flow cytometry synthetic datasets can be utilised to validate a range of automated cell identification methods, leading to enhanced confidence in the data quality of automated cell characterisations and enumerations.

Immunophenotype signatures in acute leukemias unveiled by integrative systems immunology

“…Consistent with previous analyses, we were able to effectively capture cell distribution patterns, cluster classification, variations in marker expression intensity, and differences between patient groups in our cohort. These findings emphasize the power of bioinformatics tools in analyzing flow cytometry data to differentiate between distinct AL subtypes and controls (Beyrend et al, 2018;Cheung et al, 2022;Çubukçu et al, 2023;Melsen et al, 2020;Montante & Brinkman, 2019;Saeys et al, 2016). The ability to identify leukemia-specific cellular populations and visualize differentiation trajectories can significantly optimize the diagnostic process (Ng et al, 2024;Nguyen et al, 2023;Seifert et al, 2023;Simonson et al, 2022).…”

Human leucocytes processed by fast-rate inertial microfluidics retain conventional functional characteristics

“…It should be noted that surface marker expression does not necessarily infer function and a wider assessment of function post-processing should be performed. However, these analyses do provide a good insight to viability and phenotype, with single-cell resolution, and have previously been used as a method of quality control for ATMPs [ 54 ] in accordance with GMP guidelines issued by the European Medicines Agency and the US Food and Drug Administration [ 55 ].…”

Multiomics technologies for comprehensive tumor microenvironment analysis in triple-negative breast cancer under neoadjuvant chemotherapy

“…A computational pipeline known as FlowGM, with automated identification of 24 cell types, has been demonstrated to better discriminate monocyte and dendritic cell subpopulations than the traditional gating strategy ( 71 ). Other automated cell identification tools, including PhenoGraph ( 72 ), SPADE3 ( 73 ), FlowSOM ( 74 ), SWIFT ( 75 ), t-SNE ( 76 ), and UMAP ( 77 ), distinguish cell populations from cytometry data in both unsupervised and supervised manners ( 78 , 79 ).…”