In this study we compared the performance of a conventional (liquid, CFP) versus a lyoplate-based flow cytometry platform (LFP) and the potential to integrate flow cytometry with computational data analysis to establish a robust framework to conduct biomarker discovery studies in humans.
We found that LFP has a higher sensitivity for detecting key cytokines (IFN-γ, IL-10) and activation markers (CD25, Foxp3) compared to CFP, while keeping comparable intra- and inter-assay variability.
Moreover, when computational analysis was performed by using RchyOptimix, novel immunophenotypes were identified.
Multicolour flow cytometry is becoming a preferential tool for immuno-monitoring and biomarker discovery in large human studies, thus requiring standardization of both experimental and analytical methods. However, available data refer to relatively small antibody cocktails and include the most common fluorochromes, thus only allowing the detection of a restricted set of markers 
Here, we measured for the first time 12 parameters using a LFP with and without polyclonal cell activation. Its positive performance for immunophenotyping and cytokine detection makes it a suitable alternative to CFP.
LFP has the advantage of simplifying the experimental protocol, is time saving (~3 hours in each of our experiments), and allows a medium-throughput processing of the samples, using pre-filled 96 well plates and a plate loader. Moreover, pre-formatted lyoplates, containing the same batch of reagents, can be reliably used through the entire duration of a study and across multiple centres. Therefore, LFP reduces hands-on time, while promoting automation and reagent standardization that are of primary importance in translational and clinical research studies.
Our data indicate that lyophilized reagents resulted in more powerful cell stimulation and better marker discrimination, possibly due to improved reagent stability after lyophilisation. In keeping with the increased detection of IFN-γ+, IL-10+, Foxp3+ and CD25+ cells, most of the lyophilized antibodies also resulted in increased resolution sensitivity as determined by a higher stain index (SI) on stained PBMC. Of note, tandem dyes PE-Cy5 and APC-H7 showed a decreased SI on stained PBMC compared to cells stained with liquid counterparts, indicating that lyophilisation might have a different impact on different fluorochromes. This aspect should be considered when designing the antibody cocktail to be lyophilized, and a pre-test of the lyophilisation impact onto the specific antibody-fluorochrome combinations should be performed, especially for tandem dye conjugates. If possible, choosing an antibody batch with the brightest SI on stained cells could further help balance the lyophilisation effect.
Reproducible results, a key aspect in multicenter trials, require minimal intra- and inter-assay variability. In order to reduce assay variation, we combined the use of lyoplates with strict SOPs for sample handling, rigorous instrument QC, and reproducible instrument setup. CFP and LFP showed minimal intra-assay variability, suggesting that experimental replicates are not an absolute requirement for flow cytometry analysis. Importantly, LFP allowed a more accurate detection of Tregs. This is relevant as Tregs gating is notoriously difficult and subjective 
The high dimensional data generated by multi-parameter cell analysis need to be analysed in an unsupervised, multidimensional, and fast manner to overcome the subjectivity and non-reproducibility of manual gating and analysis. In recent years, several computational tools for analysis of flow cytometry data have been developed by different research groups (see 
for a review). Two broad categories of these tools have recently been evaluated by the FlowCAP project 
:(1) Clustering algorithms for automated identification of cell populations (e.g., 
) and (2) Binary sample classification pipelines for identification of immunophenotypic differences between two groups of samples (e.g., 
). The pipeline used in this work has been designed for identification of cell populations that correlate with an external variable (e.g., a clinical outcome). Detailed descriptions are available elsewhere 
. Briefly, the pipeline can incorporate the background knowledge of the human experts into the gating process. Then, tens of thousands of immunophenotypes extracted from each sample are tested for correlation with the external variable (in this case, 6560 cell populations from every FCS file were correlated with the reagent type). Finally, the selected immunophenotypes are organized in a hierarchical structure based on their most common parent populations. These hierarchies not only provide intuitive data visualization, but also aid in adjusting the trade-off between the number of markers included in identification of a cell population of interest and the statistical significance of the correlation with the external variable. This information can also help in the use of high-dimensional datasets to guide the design of low-dimensional panels: for example a Time-of-Flight mass spectrometer (CyTOF) assay on a small dataset can analyze a large list of candidate markers, and using the hierarchies produced by this approach one can design lyoplate panels for further validation of the results.
Taken together, we propose the integration of LFP and computational analysis as a robust and standardized method for obtaining high content information on T cell proportions and functions in a medium-to-high-throughput manner. A natural application of this approach would be in the biomarker discovery arena, where the easy scalability of LFP with unbiased automated data analysis would allow the rapid and standardized screening of large human cohorts.