Beyond gating: capturing the power of flow cytometry

Computer-aided detection of rare tumor populations in flow cytometry: an example with classic Hodgkin lymphoma

OBJECTIVES

Diagnosing classical Hodgkin lymphoma (cHL) by flow cytometry (FC) relies on an observer gating rare populations of Hodgkin/Reed Sternberg (HRS) cells. Here, we apply machine-learning methods to aid in the detection of rare tumor cell populations using data derived from clinical FC analysis of cHL as a model disease.

METHODS

FC data from 144 clinical cases using a nine-color FC reagent panel were analyzed using Python 2.7 and the "scikit-learn" module.

RESULTS

Seventy-eight 50 × 50 two-dimensional histograms were generated from routine FC data and a reciprocal power function applied to favor rare events. Data were classified by support vector machine (SVM), gradient boosting, and random forest classifiers. All three classifiers showed no statistical difference in performance, with 89%-92% accuracy on cross-validation. Nearly all classifiers misclassified the same set of cases, with more false-positive than false-negative cases. Dimensionality reduction by ensemble methods selected for data points in a CD5+/ CD40+/CD64- region.

CONCLUSIONS

All classifiers provide probabilistic confidences for each result, and diagnostic cutoffs can be chosen to minimize false negatives and serve as a screening tool. Computational exclusion of manually gated HRS cells had little impact on the overall performance of selected support vectors in SVM or dimensionality reduction, suggesting that features of the immune response in cHL may dictate the method accuracy. We hypothesize there are distinct inflammatory cells that suggest cHL.

Automated analysis of flow cytometric data for measuring neutrophil CD64 expression using a multi-instrument compatible probability state model

BACKGROUND

Leuko64™ (Trillium Diagnostics) is a flow cytometric assay that measures neutrophil CD64 expression and serves as an in vitro indicator of infection/sepsis or the presence of a systemic acute inflammatory response. Leuko64 assay currently utilizes QuantiCALC, a semiautomated software that employs cluster algorithms to define cell populations. The software reduces subjective gating decisions, resulting in interanalyst variability of <5%. We evaluated a completely automated approach to measuring neutrophil CD64 expression using GemStone™ (Verity Software House) and probability state modeling (PSM).

METHODS

Four hundred and fifty-seven human blood samples were processed using the Leuko64 assay. Samples were analyzed on four different flow cytometer models: BD FACSCanto II, BD FACScan, BC Gallios/Navios, and BC FC500. A probability state model was designed to identify calibration beads and three leukocyte subpopulations based on differences in intensity levels of several parameters. PSM automatically calculates CD64 index values for each cell population using equations programmed into the model. GemStone software uses PSM that requires no operator intervention, thus totally automating data analysis and internal quality control flagging. Expert analysis with the predicate method (QuantiCALC) was performed. Interanalyst precision was evaluated for both methods of data analysis.

RESULTS

PSM with GemStone correlates well with the expert manual analysis, r(2) = 0.99675 for the neutrophil CD64 index values with no intermethod bias detected. The average interanalyst imprecision for the QuantiCALC method was 1.06% (range 0.00-7.94%), which was reduced to 0.00% with the GemStone PSM. The operator-to-operator agreement in GemStone was a perfect correlation, r(2) = 1.000.

CONCLUSION

Automated quantification of CD64 index values produced results that strongly correlate with expert analysis using a standard gate-based data analysis method. PSM successfully evaluated flow cytometric data generated by multiple instruments across multiple lots of the Leuko64 kit in all 457 cases. The probability-based method provides greater objectivity, higher data analysis speed, and allows for greater precision for in vitro diagnostic flow cytometric assays.

Deep profiling of multitube flow cytometry data

MOTIVATION

Deep profiling the phenotypic landscape of tissues using high-throughput flow cytometry (FCM) can provide important new insights into the interplay of cells in both healthy and diseased tissue. But often, especially in clinical settings, the cytometer cannot measure all the desired markers in a single aliquot. In these cases, tissue is separated into independently analysed samples, leaving a need to electronically recombine these to increase dimensionality. Nearest-neighbour (NN) based imputation fulfils this need but can produce artificial subpopulations. Clustering-based NNs can reduce these, but requires prior domain knowledge to be able to parameterize the clustering, so is unsuited to discovery settings.

RESULTS

We present flowBin, a parameterization-free method for combining multitube FCM data into a higher-dimensional form suitable for deep profiling and discovery. FlowBin allocates cells to bins defined by the common markers across tubes in a multitube experiment, then computes aggregate expression for each bin within each tube, to create a matrix of expression of all markers assayed in each tube. We show, using simulated multitube data, that flowType analysis of flowBin output reproduces the results of that same analysis on the original data for cell types of >10% abundance. We used flowBin in conjunction with classifiers to distinguish normal from cancerous cells. We used flowBin together with flowType and RchyOptimyx to profile the immunophenotypic landscape of NPM1-mutated acute myeloid leukemia, and present a series of novel cell types associated with that mutation.