A time to amaze, a time to settle down, and a time to discover

THE breakthrough manuscript of Bendall et al. in 2011 (1) demonstrated the capacity to raise the number of measurable parameters per single cells to more than 30 by using the technology of mass cytometry; being able to measure, theoretically, up to 100 parameters (2). A series of successive amazing manuscripts followed this first description of the signaling pathway in hematopoietic cells by mass cytometry. Newell et al. combined high-throughput CD8 T-cell epitope mapping to detailed phenotype analysis (3). Zunder et al. applied mass cytometry to the analysis of induced pluripotent stem cells providing a detailed view of molecular events during cellular reprogramming (4). The use of mass cytometry for the immunology field has been recently reviewed (5) showing a wide range of possible applications. However, this rapid success of mass cytometry needs to be supported by a solid dissemination on quality controls and standardization. Moreover the high complexity of the current 40 parameter data sets requires innovative analysis strategies able to render this huge amount of information interpretable for the human brain. Analysis strategies do not rely anymore on unique software and reporting system but combine multiple consecutive and parallel methods. These methods are often unsupervised in principle and agnostic in respect to the dataset and the initial scientific question. However, when the number of choices increases and the scientist make a choice, how sure are we that a complex analysis system remains “unsupervised”? Hence, it is time to settle down and build the basis for implementation of standardization of experimental methods and analysis strategies. On the experimental side, the work of Fink et al. (6) proposed a beads-based normalization system to correct for time dependent changes in instrument performances while Tricot et al. (7) described that lanthanides are detected with different sensitivities, as a function of their atomic mass and each instrument has a peculiar sensitivity pattern. These were the first two publications clearly dealing with mass cytometry experimental standardization. In the present issue (page 817), Gaudilliere et al. describe an experimental and analytical pipeline for clinical studies using mass cytometry technology. The study aims to define preterm births predictive markers in the peripheral blood of non-pregnant women with a history of preterm birth. The authors constructed the study in a hypothesis-generating setting and were able to find a putative biomarker: the capacity of classical monocyte to mount a stronger TLR4 innate response. Even if the observed differences were modest as clearly stated by the authors, the aim of the study was to generate hypothesis that can be then confirmed or confuted with less multi-parametric but more high-throughput technologies (e.g., conventional flow cytometry). Here, mass cytometry is therefore a tool to screen single cell with an unprecedented level of detail to acquire a complex information from which interesting hypothesis are generated. The article has a remarkable didactical organization to introduce the reader to the different steps used in the pipeline. Indeed, the “General consideration” introductory sections explain the reader the technical and analytical choices before entering in the detail of the methods. The parallel use of manual gating and unsupervised clustering highlight the importance to mine the dataset to uncover differences that can be otherwise overlooked;