FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2794660702> ?p ?o ?g. }

• W2794660702 endingPage "620.e5" @default.
• W2794660702 startingPage "612" @default.
• W2794660702 abstract "•Signal spillover can impact quality of mass cytometry data•Spillover can be corrected by compensation throughout the linear range of the CyTOF•Compensation enables signal correction and data structure preservation•CATALYST is a new R package and a web tool to estimate and correct for spillover The advent of mass cytometry increased the number of parameters measured at the single-cell level while decreasing the extent of crosstalk between channels relative to dye-based flow cytometry. Although reduced, spillover still exists in mass cytometry data, and minimizing its effect requires considerable expert knowledge and substantial experimental effort. Here, we describe a novel bead-based compensation workflow and R-based software that estimates and corrects for interference between channels. We performed an in-depth characterization of the spillover properties in mass cytometry, including limitations defined by the linear range of the mass cytometer and the reproducibility of the spillover over time and across machines. We demonstrated the utility of our method in suspension and imaging mass cytometry. To conclude, our approach greatly simplifies the development of new antibody panels, increases flexibility for antibody-metal pairing, opens the way to using less pure isotopes, and improves overall data quality, thereby reducing the risk of reporting cell phenotype artifacts. The advent of mass cytometry increased the number of parameters measured at the single-cell level while decreasing the extent of crosstalk between channels relative to dye-based flow cytometry. Although reduced, spillover still exists in mass cytometry data, and minimizing its effect requires considerable expert knowledge and substantial experimental effort. Here, we describe a novel bead-based compensation workflow and R-based software that estimates and corrects for interference between channels. We performed an in-depth characterization of the spillover properties in mass cytometry, including limitations defined by the linear range of the mass cytometer and the reproducibility of the spillover over time and across machines. We demonstrated the utility of our method in suspension and imaging mass cytometry. To conclude, our approach greatly simplifies the development of new antibody panels, increases flexibility for antibody-metal pairing, opens the way to using less pure isotopes, and improves overall data quality, thereby reducing the risk of reporting cell phenotype artifacts. High-dimensional, single-cell flow cytometry has been broadly adopted by researchers and clinicians to analyze complex biological samples (Behbehani et al., 2012Behbehani G.K. Bendall S.C. Clutter M.R. Fantl W.J. Nolan G.P. Single-cell mass cytometry adapted to measurements of the cell cycle.Cytometry A. 2012; 81: 552-566Crossref PubMed Scopus (164) Google Scholar, Chevrier et al., 2017Chevrier S. Levine J.H. Zanotelli V.R.T. Silina K. Schulz D. Bacac M. Ries C.H. Ailles L. Jewett M.A.S. Moch H. et al.An immune atlas of clear cell renal cell carcinoma.Cell. 2017; 169: 736-749.e18Abstract Full Text Full Text PDF PubMed Scopus (568) Google Scholar, Levine et al., 2015Levine J.H. Simonds E.F. Bendall S.C. Davis K.L. Amir E.D. Tadmor M.D. Litvin O. Fienberg H.G. Jager A. Zunder E.R. et al.Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis.Cell. 2015; 162: 184-197Abstract Full Text Full Text PDF PubMed Scopus (986) Google Scholar, Perfetto et al., 2004Perfetto S.P. Chattopadhyay P.K. Roederer M. Seventeen-colour flow cytometry: unravelling the immune system.Nat. Rev. Immunol. 2004; 4: 648-655Crossref PubMed Scopus (821) Google Scholar). Fluorescence-activated cell sorting (FACS) has dominated this field for decades, and, with the constant improvement of probes and laser systems, 18-color FACS experiments are now routine, and 30-color cytometers have recently become commercially available (Chattopadhyay and Roederer, 2012Chattopadhyay P.K. Roederer M. Cytometry: today’s technology and tomorrow’s horizons.Methods. 2012; 57: 251-258Crossref PubMed Scopus (97) Google Scholar). Due to the overlapping excitation and emission spectra of the fluorescent dyes, signals are measured not only in the primary channel, but also in neighboring channels. This spillover is correlated with the original signal in an approximately linear manner and can be corrected via a process called compensation (Bagwell and Adams, 1993Bagwell C.B. Adams E.G. Fluorescence spectral overlap compensation for any number of flow cytometry parameters.Ann. N. Y. Acad. Sci. 1993; 677: 167-184Crossref PubMed Scopus (107) Google Scholar). As the number of parameters measured increases, however, it becomes more difficult to optimize artifact-free staining panels, mostly due to the spreading error affecting channels to different extents upon compensation, which complicates the detection of proteins of low abundance (Chattopadhyay and Roederer, 2012Chattopadhyay P.K. Roederer M. Cytometry: today’s technology and tomorrow’s horizons.Methods. 2012; 57: 251-258Crossref PubMed Scopus (97) Google Scholar). Mass cytometry, which uses metal isotopes as reporter to label antibodies, allows analysis of at least 40 parameters simultaneously (Bandura et al., 2009Bandura D.R. Baranov V.I. Ornatsky O.I. Antonov A. Kinach R. Lou X. Pavlov S. Vorobiev S. Dick J.E. Tanner S.D. Mass cytometry: technique for real time single cell multitarget immunoassay based on inductively coupled plasma time-of-flight mass spectrometry.Anal. Chem. 2009; 81: 6813-6822Crossref PubMed Scopus (882) Google Scholar, Bendall et al., 2011Bendall S.C. Simonds E.F. Qiu P. Amir A.D. Krutzik P.O. Finck R. Bruggner R.V. Melamed R. Trejo A. Ornatsky O.I. et al.Single- cell mass cytometry of differential immune and drug responses across a hu- man hematopoietic continuum.Science. 2011; 332: 687-696Crossref PubMed Scopus (1680) Google Scholar, Chattopadhyay and Roederer, 2015Chattopadhyay P.K. Roederer M. A mine is a terrible thing to waste: high content, single cell technologies for comprehensive immune analysis.Am. J. Transplant. 2015; 15: 1155-1161Crossref PubMed Scopus (35) Google Scholar). This technology has recently been exploited for imaging by coupling a laser ablation system to a mass cytometer (Bodenmiller, 2016Bodenmiller B. Multiplexed epitope-based tissue imaging for discovery and healthcare applications.Cell Syst. 2016; 2: 225-238Abstract Full Text Full Text PDF PubMed Scopus (149) Google Scholar, Giesen et al., 2014Giesen C. Wang H.A.O. Schapiro D. Zivanovic N. Jacobs A. Hattendorf B. Schüffler P.J. Grolimund D. Buhmann J.M. Brandt S. et al.Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry.Nat. Methods. 2014; 11: 417-422Crossref PubMed Scopus (986) Google Scholar, Angelo et al., 2014Angelo M. Bendall S.C. Finck R. Hale M.B. Hitzman C. Borowsky A.D. Levenson R.M. Lowe J.B. Liu S.D. Zhao S. et al.Multiplexed ion beam imaging of human breast tumors.Nat. Med. 2014; 20: 436-442Crossref PubMed Scopus (607) Google Scholar). Imaging mass cytometry (IMC) enables the analysis of tissue sections stained with metal-tagged antibodies to generate highly multiplexed images at subcellular resolution (Bodenmiller, 2016Bodenmiller B. Multiplexed epitope-based tissue imaging for discovery and healthcare applications.Cell Syst. 2016; 2: 225-238Abstract Full Text Full Text PDF PubMed Scopus (149) Google Scholar, Giesen et al., 2014Giesen C. Wang H.A.O. Schapiro D. Zivanovic N. Jacobs A. Hattendorf B. Schüffler P.J. Grolimund D. Buhmann J.M. Brandt S. et al.Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry.Nat. Methods. 2014; 11: 417-422Crossref PubMed Scopus (986) Google Scholar). Although the amount of spillover observed in mass cytometry is generally small, spillover can considerably complicate interpretation of data and potentially lead to false conclusions. For example, signal crosstalk can result in incorrect identification of cells as expressing an intermediate level of a marker (Takahashi et al., 2016Takahashi C. Au-Yeung A. Fuh F. Ramirez-Montagut T. Bolen C. Mathews W. O’Gorman W.E. Mass cytometry panel optimization through the designed distribution of signal interference.Cytometry A. 2016; 91: 39-47Crossref PubMed Scopus (48) Google Scholar). In experiments conducted to date, the effects of spillover have been minimized by selecting only highly pure isotopes and by carefully designing antibody panels to optimize the signal to background ratio in each channel (Takahashi et al., 2016Takahashi C. Au-Yeung A. Fuh F. Ramirez-Montagut T. Bolen C. Mathews W. O’Gorman W.E. Mass cytometry panel optimization through the designed distribution of signal interference.Cytometry A. 2016; 91: 39-47Crossref PubMed Scopus (48) Google Scholar). Generating a low crosstalk antibody panel is complex and time consuming, however. It requires that the approximate antigen abundance is known for each marker used in the panel, which is not possible in many types of experiments. Further, with a purely experimental approach to avoid spillover, antibody-isotope conjugates are not easily transferable between panels. As spillover is proportional to the originating signal, it can be reduced by decreasing antibody concentrations, but this also reduces the signal-to-noise ratio, which limits its application. In practice, the above-mentioned strategies are not sufficient to completely prevent crosstalk between channels as shown in a recent study in which data from spillover-affected channels were excluded to avoid potentially misleading conclusions (Lun et al., 2017Lun X.K. Zanotelli V.R. Wade J.D. Schapiro D. Tognetti M. Dobberstein N. Bodenmiller B. Influence of node abundance on signaling network state and dynamics analyzed by mass cytometry.Nat. Biotechnol. 2017; 35: 164-172Crossref PubMed Scopus (25) Google Scholar). Spillover-related issues have not yet been reported in IMC, but since the source and the measurement of metal signal in suspension mass cytometry and IMC are identical, both systems are expected to be affected in a similar manner. Here we present a comprehensive workflow to estimate and systematically correct for signal spillover across all the channels used in a given mass cytometry experiment. Polystyrene capture beads were single stained with each antibody used in the experiment. To increase the throughput, the beads were then pooled and analyzed simultaneously in the mass cytometer. Mixing is critical to efficiency as it allows assessment of spillover in channels within minutes. The CATALYST R/Bioconductor package and an interactive Shiny-based web application were developed to accurately deconvolute the different bead populations, estimate spillover signal in all channels, and compensate the data. We demonstrate the utility of the approach in correction of signal interference in suspension mass cytometry and IMC experiments. Our approach will greatly facilitate the development of antibody panels, increase the flexibility of antibody-metal pairing, increase the number of usable isotopes, and enable generation of high-quality data devoid of spillover artifacts on samples with unknown and likely variable levels of antigen. Fluorescent flow cytometry is affected by signal interference between channels. Since spillover signal is a defined fraction of the source signal, it can be corrected mathematically (Bagwell and Adams, 1993Bagwell C.B. Adams E.G. Fluorescence spectral overlap compensation for any number of flow cytometry parameters.Ann. N. Y. Acad. Sci. 1993; 677: 167-184Crossref PubMed Scopus (107) Google Scholar, Loken et al., 1977Loken M.R. Parks D.R. Herzenberg L.A. Two-color immunofluorescence using a fluorescence-activated cell sorter.J. Histochem. Cytochem. 1977; 25: 899-907Crossref PubMed Scopus (147) Google Scholar). In mass cytometry, the interference between channels is reduced but is still present due to instrument properties (abundance sensitivity), isotopic impurities, and oxidation (Figure 1A). To determine whether channel crosstalk observed in mass cytometry can be corrected in a manner similar to the one used for flow cytometry, we first determined whether the crosstalk in mass cytometry experiments is linear. We stained peripheral blood mononuclear cells (PBMCs) with anti-CD44 conjugated to 143Nd using antibody concentrations ranging from 0.01 to 1 μg/mL (Figure 1B). As expected, signal was observed in other mass channels including −1 (142Nd), +1 (144Nd), +2 (145Nd), +3 (146Nd), and +16 (due to the oxidation product 143Nd16O measured in 159Tb). The signal in the source and in the spillover channels could be fit by a linear model with a coefficient of determination (R2) greater than 0.99 in all cases (Figure 1C, left panels). Moreover, we showed that a signal over 200 counts was sufficient to provide an accurate estimate for spillover as low as 1%, calculated as the ratio of the spillover signal to the main signal for each concentration (Figure 1C, right panels). Applying these spillover coefficients on the single-stained cells removed the spillover (Figure 1D, middle panels). However, this strategy substantially modified the structure of the data by introducing artificial negative values (Figure 1D, compare orange and blue boxes), which specifically influenced channels strongly affected by spillover. Negative ion counts are not present in uncompensated mass cytometry data, and, more importantly, data with negative values require different treatment than strictly non-negative abundance data. A recent study aimed at unmixing signals in multispectral fluorescent flow cytometry made similar observations and suggested use of approaches that specifically incorporate a non-negativity constraint such as the non-negative least-squares (NNLS) approach (Novo et al., 2014Novo D. Grégori G. Bartek R. Generalized unmixing model for multispectral flow cytometry utilizing nonsquare compensation matrices.Cytometry A. 2014; 83: 508-520Google Scholar). This method calculates the optimal non-negative solution for the compensation problem using the least-squares criterion. Applied to our data, the NNLS approach removed the spillover without changing the data structure, making empty but spillover-affected channels look similar to empty channels not affected by spillover (Figure 1D, compare green and blue boxes). Taken together, our data show that spillover in mass cytometry is linear and can be corrected while preserving the data structure using the NNLS approach. Inspired by methods used in flow cytometry, in which controls stained with single antibodies are used to estimate signal crosstalk, we developed an approach to systematically correct for signal interference in mass cytometry experiments. A 36-antibody panel was designed to detect the main immune cell populations in PBMCs (Table S1). This panel was not optimized to avoid spillover effects and contained identical antibodies in different mass channels to facilitate the identification of spillover artifacts. In parallel to multiplexed sample staining, control samples stained with individual antibodies were generated by staining polystyrene antibody-capture beads (Figure 1E). After staining, beads were pooled and run as a single sample in the mass cytometer. To apply our approach for semi-automatic spillover correction in mass cytometry, we created an R/Bioconductor package, CATALYST, and a web application (Figures S1A and S1B). In the first step, the FCS file containing data on the bead sample is deconvoluted to identify the individual single-antibody-positive bead populations using a new R implementation of the debarcoding algorithm from Zunder et al., 2015Zunder E.R. Finck R. Behbehani G.K. Amir E.-A.D. Krishnaswamy S. Gonzalez V.D. Lorang C.G. Bjornson Z. Spitzer M.H. Bodenmiller B. et al.Palladium-based mass tag cell barcoding with a doublet-filtering scheme and single-cell deconvolution algorithm.Nat. Protoc. 2015; 10: 316-333Crossref PubMed Scopus (312) Google Scholar. Each bead is assigned to a specific population based on the dominant signal, and the purity of the bead populations is further increased by automatically applying estimated sample-specific cutoffs (Figure S1C). Upon debarcoding, the purity of the bead populations was assessed to ensure that no beads had been wrongly assigned and that no antibody exchange had occurred during the bead mixing, which could influence the spillover estimate (Figure S2A). In a second step, the spillover matrix is calculated based on the spillover observed for single-stained populations. Due to the mass cytometry data structure, characterized by an absence of negative values and a low background, we observed that spillover estimation was more accurate when the spillover was assessed at the single-bead level rather than at the bead-population level (Figures S2B–S2D; see the STAR Methods for details). By default, the method only takes into account interference between channels expected to interact based on abundance sensitivity, metal impurity, and oxidation (Figures S2E and S2F), but also allows the user to check for unexpected spillover. In a final step, the compensation matrix from the solved linear system (NNLS or “classical”) is applied to the bead and cell samples to remove interfering signal. This workflow provides a fully integrated and easy to use experimental and computational solution for compensation of mass cytometry spillover. The spillover matrix generated by our bead approach revealed that the total amount of spillover originating from a single channel ranged from 0% for 165Ho to over 8% for 148Nd, oxidation ranged from 0% to 2%, and spillover due to mixed effects of impurity and abundance sensitivity may reach 4% (Figure 2A). Signal interference due to abundance sensitivity alone was virtually absent on the machine used. To assess the stability of the spillover matrix over time and instruments, we collected data on single-stained beads on different mass cytometers over months (Figure S3). Although the spillover matrix was stable over months, our results showed that for optimal compensation the spillover matrix should be acquired simultaneously with the sample of interest. As expected, the application of the spillover matrix to beads stained with individual antibodies revealed virtually perfect compensation using both traditional and NNLS approaches (Figures 2B and 2C). When this matrix was applied to the multiplexed-stained cell samples, the spillover was also corrected for, but traditional compensation systematically overcompensated the data (Figures 2B and 2C). One possible explanation for the difference observed in spillover between single-stained beads and multiplexed-stained cells might be the difference in total ion load, with high loads leading to detector saturation effects. Indeed, we found that an increased amount of barcoding, simulating higher ion loads, was associated with a progressive decrease of spillover, both in terms of percentage and absolute count (Figures 2D and 2E). For spillover below two counts, the signal interference was completely abolished (Figure 2E). Moreover, above 5,000 dual counts in a given channel, the linear relationship was progressively lost (Figure 2F). Together, this set of data revealed some limitations of spillover correction due to the physical properties of mass cytometry measurement and showed that maintaining the signal within the linear range of the instrument and using the NNLS compensation addresses these issues. Analyzing the PBMCs stained with the 36-antibody panel using the dimensionality reduction algorithm t-SNE (Amir et al., 2013Amir E.D. Davis K.L. Tadmor M.D. Simonds E.F. Levine J.H. Bendall S.C. Shenfeld D.K. Krishnaswamy S. Nolan G.P. Pe’er D. viSNE enables visualization of high dimensional single-cell data and reveals phenotypic heterogeneity of leukemia.Nat. Biotechnol. 2013; 31: 545-552Crossref PubMed Scopus (1102) Google Scholar, van der Maaten and Hinton, 2008van Der Maaten L. Hinton G. Visualizing data using t-SNE.J. Mach. Learn. Res. 2008; 9: 2579-2605Google Scholar) enabled us to identify the main immune cell populations based on individual marker expression (Figure S4A). The proteins CD3, CD8, and HLA-DR were each detected with antibodies conjugated to two different metal labels. In uncompensated data, we observed different signal profiles that depended on the metal isotope used to label the antibody (Figure 3A, left panel). After compensation, the signals observed for the same antibodies conjugated with different metal isotopes were virtually identical (Figures 3A, right panel and S4B). This was further demonstrated by displaying the same relationships on scatterplots, which highlighted how compensation simultaneously removed artefactual signal and reconstituted the data structure observed in channel stained with the same antibody but not affected by spillover (Figure 3B). Thus, compensation removes artifacts and therefore prevents data misinterpretation. Applying the PhenoGraph clustering algorithm (Levine et al., 2015Levine J.H. Simonds E.F. Bendall S.C. Davis K.L. Amir E.D. Tadmor M.D. Litvin O. Fienberg H.G. Jager A. Zunder E.R. et al.Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis.Cell. 2015; 162: 184-197Abstract Full Text Full Text PDF PubMed Scopus (986) Google Scholar) to our dataset led to the identification of 20 PBMC subsets (Figure S4C). Comparison of heatmaps of signals in uncompensated versus compensated data highlighted how marker expression signatures of the different clusters can be misinterpreted without spillover correction (Figure 3C). Lack of compensation caused several clusters to be wrongly identified as having intermediate abundances of certain antigens even though the signal was actually due to channel crosstalk. In particular, an intermediate level of CD3-173Yb was observed on all the non-T cell subsets (Figure 3C). Further, most T cell and natural killer cell subsets were wrongly identified as expressing intermediate levels of HLA-DR-171Yb. Artifacts caused by crosstalk were particularly strong in channels 154, 158, 161, 163, 168, 171, 173, and 174. Characterization of newly identified clusters or signaling network inference often involves the systematic correlation analysis of markers at the single-cell level to identify co-regulated proteins or genes, and this approach can be strongly affected by channel interference. Analysis of marker correlations within each cluster before and after compensation systematically reduced spurious correlations (Figures 3D and 3E). A systematic analysis over all the clusters showed that, in our dataset, between 25% and 45% of the significant marker correlations within clusters were actually due to spillover (Figures 3D and 3E, Spearman correlation, p < 0.005). Collectively, this set of analyses showed that spillover can be responsible for various artifacts, which were removed with our compensation approach. In IMC, tissue stained with metal-tagged antibodies is ablated with a laser, and the tissue aerosol is analyzed in a mass cytometer (Giesen et al., 2014Giesen C. Wang H.A.O. Schapiro D. Zivanovic N. Jacobs A. Hattendorf B. Schüffler P.J. Grolimund D. Buhmann J.M. Brandt S. et al.Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry.Nat. Methods. 2014; 11: 417-422Crossref PubMed Scopus (986) Google Scholar). Images generated with the IMC system provide subcellular resolution and are high dimensional; information has been collected from 32 different channels (Bodenmiller, 2016Bodenmiller B. Multiplexed epitope-based tissue imaging for discovery and healthcare applications.Cell Syst. 2016; 2: 225-238Abstract Full Text Full Text PDF PubMed Scopus (149) Google Scholar, Giesen et al., 2014Giesen C. Wang H.A.O. Schapiro D. Zivanovic N. Jacobs A. Hattendorf B. Schüffler P.J. Grolimund D. Buhmann J.M. Brandt S. et al.Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry.Nat. Methods. 2014; 11: 417-422Crossref PubMed Scopus (986) Google Scholar). To determine how signal interference affects IMC measurements, metal isotopes were arrayed on a slide and measured by IMC. Using this approach, we demonstrated that a linear relationship exists between the original signal and the interfering signals (Figure 4A). This indicated that spillover in IMC could be corrected using the bead-based compensation approach applied to suspension mass cytometry. We used the CATALYST package to calculate a spillover matrix based on the pixel values of the individually spotted heavy metals (Figure S5A). Comparing individual spillover values obtained in suspension and IMC, we found that spillover due to abundance sensitivity and impurities were in the same range for all metals except for 148Nd and 176Yb, which came from different isotope batches for the IMC experiment than those used for the suspension analysis (Figures S5B–S5D). Values observed for oxidation in the M+16 channel were systematically lower in IMC than in suspension mass cytometry. This was expected given that the tissue aerosol is transported in an argon and helium gas stream and no water is used for sample introduction, thus much less oxygen is present in the plasma of the mass cytometer to generate oxides (Figure S5E). Based on this spillover matrix, a breast cancer tissue section imaged by IMC was compensated at the pixel level using a custom written CellProfiler module (Carpenter et al., 2006Carpenter A.E. Jones T.R. Lamprecht M.R. Clarke C. Kang I.H. Friman O. Guertin D.A Chang J.H. Lindquist R.A Moffat J. et al.CellProfiler: image analysis software for identifying and quantifying cell phenotypes.Genome Biol. 2006; 7: R100Crossref PubMed Scopus (3350) Google Scholar). This compensation approach specifically removed the low signal due to spillover (Figure 4B). The carbonic anhydrase antibody, which was labeled with 166Er, showed a predominantly membranous signal. No antibodies were labeled with 167Er (the +1 channel) to enable assessment of the spillover. In the uncompensated data, there was a perfect, but lower intensity image of the 166Er channel in the 167Er channel. Thus, without compensation, the channel 167Er is not suitable for detection of a low-level marker. 168Er was used to label an antibody against Ki67, a protein tightly regulated during cell-cycle progression. Even though the spillover from 166Er into 168Er is only estimated to be 0.2%, due to the low background in IMC, the carbonic anhydrase signal was still clearly visible in the Ki67 channel. This could lead to the misinterpretation that Ki67 is localized in the cytoplasm and on the membrane. Upon compensation, the shadow images of the 166Er channel in the 167Er and 168Er channels were removed (Figure 4B). IMC images are often segmented to identify individual cells in the images enabling single-cell data analysis (Giesen et al., 2014Giesen C. Wang H.A.O. Schapiro D. Zivanovic N. Jacobs A. Hattendorf B. Schüffler P.J. Grolimund D. Buhmann J.M. Brandt S. et al.Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry.Nat. Methods. 2014; 11: 417-422Crossref PubMed Scopus (986) Google Scholar, Schapiro et al., 2017Schapiro D. Jackson H.W. Raghuraman S. Fischer J.R. Zanotelli V.R.T. Schulz D. Giesen C. Catena R. Varga Z. Bodenmiller B. histoCAT: analysis of cell phenotypes and interactions in multiplex image cytometry data.Nat. Methods. 2017; 14: 873-876Crossref PubMed Scopus (248) Google Scholar). After segmentation, the single-cell mask mean signal intensities for the channels of interest were calculated using a customized CellProfiler module and subsequently analyzed and compensated in R using the CATALYST package (Figure 4C). A scatterplot of this image clearly shows the spillover artifact and how our compensation approach applied to the single-cell data largely removed it (Figure 4D). Together, this set of data indicates that spillover observed in IMC data can be, and should be, corrected using the compensation approach developed for correction of suspension mass cytometry data. Relative to fluorescence leakage in flow cytometry, channel interference is considerably reduced in mass cytometry, but it is not absent. Two reports have highlighted the challenges posed by spillover to mass cytometry data analysis and interpretation (Lun et al., 2017Lun X.K. Zanotelli V.R. Wade J.D. Schapiro D. Tognetti M. Dobberstein N. Bodenmiller B. Influence of node abundance on signaling network state and dynamics analyzed by mass cytometry.Nat. Biotechnol. 2017; 35: 164-172Crossref PubMed Scopus (25) Google Scholar, Takahashi et al., 2016Takahashi C. Au-Yeung A. Fuh F. Ramirez-Montagut T. Bolen C. Mathews W. O’Gorman W.E. Mass cytometry panel optimization through the designed distribution of signal interference.Cytometry A. 2016; 91: 39-47Crossref PubMed Scopus (48) Google Scholar). Although issues related to channel interference in IMC analyses have not yet been reported, our observations show that imaging and suspension mass cytometry are similarly affected by signal interference between channels. The main issue is the virtually absent background signal in mass cytometry, which enables the reliable detection of signals at low counts (∼10 counts). Given the dynamic range over four orders of magnitude in mass cytometry, even a few percent signal spillover from a high ion count channel into a low ion count channel can easily result in difficulties in data interpretation. Moreover, high-dimensional mass cytometry data are commonly analyzed using unsupervised approaches, which present many advantages but involve a risk of misinterpretation of the data due to such artifacts. Currently, mass cytometry is transitioning from an emerging to a well-established tec" @default.
• W2794660702 created "2018-04-06" @default.
• W2794660702 creator A5005136695 @default.
• W2794660702 creator A5010642463 @default.
• W2794660702 creator A5021475120 @default.
• W2794660702 creator A5025531524 @default.
• W2794660702 creator A5029455059 @default.
• W2794660702 creator A5069576688 @default.
• W2794660702 date "2018-05-01" @default.
• W2794660702 modified "2023-10-16" @default.
• W2794660702 title "Compensation of Signal Spillover in Suspension and Imaging Mass Cytometry" @default.
• W2794660702 cites W1495845290 @default.
• W2794660702 cites W1537768706 @default.
• W2794660702 cites W1631320694 @default.
• W2794660702 cites W1767380958 @default.
• W2794660702 cites W1970170720 @default.
• W2794660702 cites W1982729887 @default.
• W2794660702 cites W2016322906 @default.
• W2794660702 cites W2020487355 @default.
• W2794660702 cites W2024626836 @default.
• W2794660702 cites W2031970585 @default.
• W2794660702 cites W2038254572 @default.
• W2794660702 cites W2038426170 @default.
• W2794660702 cites W2043565262 @default.
• W2794660702 cites W2053129129 @default.
• W2794660702 cites W2065666599 @default.
• W2794660702 cites W2107554012 @default.
• W2794660702 cites W2121469024 @default.
• W2794660702 cites W2150756255 @default.
• W2794660702 cites W2158333823 @default.
• W2794660702 cites W2344591493 @default.
• W2794660702 cites W2460271701 @default.
• W2794660702 cites W2522830304 @default.
• W2794660702 cites W2574795978 @default.
• W2794660702 cites W2611919322 @default.
• W2794660702 cites W2743344672 @default.
• W2794660702 cites W4254020410 @default.
• W2794660702 doi "https://doi.org/10.1016/j.cels.2018.02.010" @default.
• W2794660702 hasPubMedCentralId "https://www.ncbi.nlm.nih.gov/pmc/articles/5981006" @default.
• W2794660702 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/29605184" @default.
• W2794660702 hasPublicationYear "2018" @default.
• W2794660702 type Work @default.
• W2794660702 sameAs 2794660702 @default.
• W2794660702 citedByCount "240" @default.
• W2794660702 countsByYear W27946607022018 @default.
• W2794660702 countsByYear W27946607022019 @default.
• W2794660702 countsByYear W27946607022020 @default.
• W2794660702 countsByYear W27946607022021 @default.
• W2794660702 countsByYear W27946607022022 @default.
• W2794660702 countsByYear W27946607022023 @default.
• W2794660702 crossrefType "journal-article" @default.
• W2794660702 hasAuthorship W2794660702A5005136695 @default.
• W2794660702 hasAuthorship W2794660702A5010642463 @default.
• W2794660702 hasAuthorship W2794660702A5021475120 @default.
• W2794660702 hasAuthorship W2794660702A5025531524 @default.
• W2794660702 hasAuthorship W2794660702A5029455059 @default.
• W2794660702 hasAuthorship W2794660702A5069576688 @default.
• W2794660702 hasBestOaLocation W27946607021 @default.
• W2794660702 hasConcept C104317684 @default.
• W2794660702 hasConcept C105341887 @default.
• W2794660702 hasConcept C11171543 @default.
• W2794660702 hasConcept C127716648 @default.
• W2794660702 hasConcept C15744967 @default.
• W2794660702 hasConcept C162324750 @default.
• W2794660702 hasConcept C175444787 @default.
• W2794660702 hasConcept C185592680 @default.
• W2794660702 hasConcept C186060115 @default.
• W2794660702 hasConcept C199360897 @default.
• W2794660702 hasConcept C202444582 @default.
• W2794660702 hasConcept C2778015335 @default.
• W2794660702 hasConcept C2779843651 @default.
• W2794660702 hasConcept C2780023022 @default.
• W2794660702 hasConcept C33923547 @default.
• W2794660702 hasConcept C41008148 @default.
• W2794660702 hasConcept C55493867 @default.
• W2794660702 hasConcept C55527203 @default.
• W2794660702 hasConcept C5961521 @default.
• W2794660702 hasConcept C86803240 @default.
• W2794660702 hasConceptScore W2794660702C104317684 @default.
• W2794660702 hasConceptScore W2794660702C105341887 @default.
• W2794660702 hasConceptScore W2794660702C11171543 @default.
• W2794660702 hasConceptScore W2794660702C127716648 @default.
• W2794660702 hasConceptScore W2794660702C15744967 @default.
• W2794660702 hasConceptScore W2794660702C162324750 @default.
• W2794660702 hasConceptScore W2794660702C175444787 @default.
• W2794660702 hasConceptScore W2794660702C185592680 @default.
• W2794660702 hasConceptScore W2794660702C186060115 @default.
• W2794660702 hasConceptScore W2794660702C199360897 @default.
• W2794660702 hasConceptScore W2794660702C202444582 @default.
• W2794660702 hasConceptScore W2794660702C2778015335 @default.
• W2794660702 hasConceptScore W2794660702C2779843651 @default.
• W2794660702 hasConceptScore W2794660702C2780023022 @default.
• W2794660702 hasConceptScore W2794660702C33923547 @default.
• W2794660702 hasConceptScore W2794660702C41008148 @default.
• W2794660702 hasConceptScore W2794660702C55493867 @default.
• W2794660702 hasConceptScore W2794660702C55527203 @default.
• W2794660702 hasConceptScore W2794660702C5961521 @default.
• W2794660702 hasConceptScore W2794660702C86803240 @default.

• next