Search

We present distinct, a general method for differential analysis of full distributions that is well suited to applications on single-cell data, such as single-cell RNA sequencing and high-dimensional flow or mass cytometry data. High-throughput single-cell data reveal an unprecedented view of cell identity and allow complex variations between conditions to be discovered; nonetheless, most methods for differential expression target differences in the mean and struggle to identify changes where the mean is only marginally affected. distinct is based on a hierarchical non-parametric permutation ap- proach and, by comparing empirical cumulative distribution functions, iden- tifies both differential patterns involving changes in the mean, as well as more subtle variations that do not involve the mean. We performed extensive bench- marks across both simulated and experimental datasets from single-cell RNA sequencing and mass cytometry data, where distinct shows favourable per- formance, identifies more differential patterns than competitors, and displays good control of false positive and false discovery rates. distinct is available as a Bioconductor R package.

BackgroundSingle-cell gene expression profiling provides unique opportunities to understand tumor heterogeneity and the tumor microenvironment. Because of cost and feasibility, profiling bulk tumors remains the primary population-scale analytical strategy. Many algorithms can deconvolve these tumors using single-cell profiles to infer their composition. While experimental choices do not change the true underlying composition of the tumor, they can affect the measurements produced by the assay.ResultsWe generated a dataset of high-grade serous ovarian tumors with paired expression profiles from using multiple strategies to examine the extent to which experimental factors impact the results of downstream tumor deconvolution methods. We find that pooling samples for single-cell sequencing and subsequent demultiplexing has a minimal effect. We identify dissociation-induced differences that affect cell composition, leading to changes that may compromise the assumptions underlying some deconvolution algorithms. We also observe differences across mRNA enrichment methods that introduce additional discrepancies between the two data types. We also find that experimental factors change cell composition estimates and that the impact differs by method.ConclusionsPrevious benchmarks of deconvolution methods have largely ignored experimental factors. We find that methods vary in their robustness to experimental factors. We provide recommendations for methods developers seeking to produce the next generation of deconvolution approaches and for scientists designing experiments using deconvolution to study tumor heterogeneity.

High-dimensional cytometry experiments measuring 20–50 cellular markers have become routine in many laboratories. The increased complexity of these datasets requires added rigor during the experimental planning and the subsequent manual and computational data analysis to avoid artefacts and misinterpretation of results. Here we discuss pitfalls frequently encountered during high-dimensional cytometry data analysis and aim to provide a basic framework and recommendations for reporting and analyzing these datasets.

Norepinephrine (NE) neurons in the locus coeruleus (LC) project widely throughout the central nervous system, playing critical roles in arousal and mood, as well as various components of cognition including attention, learning, and memory. The LC-NE system is also implicated in multiple neurological and neuropsychiatric disorders. Importantly, LC-NE neurons are highly sensitive to degeneration in both Alzheimer’s and Parkinson’s disease. Despite the clinical importance of the brain region and the prominent role of LC-NE neurons in a variety of brain and behavioral functions, a detailed molecular characterization of the LC is lacking. Here, we used a combination of spatially-resolved transcriptomics and single-nucleus RNA-sequencing to characterize the molecular landscape of the LC region and the transcriptomic profile of LC-NE neurons in the human brain. We provide a freely accessible resource of these data in web-accessible formats.

Background Spatially-resolved transcriptomics has now enabled the quantification of high-throughput and transcriptome-wide gene expression in intact tissue while also retaining the spatial coordinates. Incorporating the precise spatial mapping of gene activity advances our understanding of intact tissue-specific biological processes. In order to interpret these novel spatial data types, interactive visualization tools are necessary. Results We describe spatialLIBD, an R/Bioconductor package to interactively explore spatially-resolved transcriptomics data generated with the 10x Genomics Visium platform. The package contains functions to interactively access, visualize, and inspect the observed spatial gene expression data and data-driven clusters identified with supervised or unsupervised analyses, either on the user’s computer or through a web application. Conclusions spatialLIBD is available at https://bioconductor.org/packages/spatialLIBD. It is fully compatible with SpatialExperiment and the Bioconductor ecosystem. Its functionality facilitates analyzing and interactively exploring spatially-resolved data from the Visium platform.

Feature selection to identify spatially variable genes or other biologically informative genes is a key step during analyses of spatially-resolved transcriptomics data. Here, we propose nnSVG, a scalable approach to identify spatially variable genes based on nearest-neighbor Gaussian processes. Our method (i) identifies genes that vary in expression continuously across the entire tissue or withina prioridefined spatial domains, (ii) uses gene-specific estimates of length scale parameters within the Gaussian process models, and (iii) scales linearly with the number of spatial locations. We demonstrate the performance of our method using experimental data from several technological platforms and simulations. A software implementation is available athttps://bioconductor.org/packages/nnSVG.

Background: The role of inflammatory immune responses in colorectal cancer (CRC) development and response to therapy is a matter of intense debate. While inflammation is a known driver of CRC, inflammatory immune infiltrates are a positive prognostic factor in CRC as inflammation can enhance response to immune checkpoint blockade (ICB) therapy. Yet, the majority of CRC are poorly infiltrated and even in presence of an inflammatory infiltrate half of these cancers become refractory to ICB. Therefore, there is a need to identify mechanisms to establish therapy supportive inflammation. We have previously reported that, irrespective of their MSI-H or MSS status, about 10% of patients with CRC down-regulate the receptor for the complement anaphylatoxin C3a (C3aR) via methylation-dependent mechanisms. The complement system is a well characterized first line of defense against pathogens and a central component of the immune response. Emerging evidence suggests that complement C3aR may play a role in intestinal homeostasis. However, to date, it is unknown how the C3a/C3aR axis can affect CRC development and influence response to therapy with ICB. Methods: We used here a comprehensive approach encompassing analysis of publicly available human CRC human datasets, inflammation-driven and newly generated spontaneous mouse models of CRC, and multi-platform high dimensional analysis of immune responses and microbiota sequencing, RNASeq and mass cytometry. Results: Using our novel mouse model of CRC (APCMin/+/C3aR-/-) we showed that loss of C3aR resulted in enhanced immune infiltration in typically “cold” tumors and changes in the intestinal microbiota. Similarly, we found that down-regulation of the complement C3aR in patients correlates with changes in specific tumor-infiltrating immune cells. Notably, when treated with a-PD1, APCMin/+/C3aR-/- mice showed a significant reduction in their tumor multiplicity and tumor load compared to untreated mice, suggesting that the loss of C3aR unleashes in the colon a microbiota-mediated favorable inflammatory infiltrate that can be exploited by the immunotherapy to reduce both de novo tumor development and the growth of already existing tumors. Conclusions: Our findings reveal that C3aR may act as a previously unrecognized checkpoint to enhance antitumor immunity in CRC. Our results suggest that C3aR may be exploited to overcome ICB resistance in a larger group of CRC patients. Citation Format: Silvia Guglietta, Lukas M. Weber, Bruno Fosso, Marinella Marzano, Gary Hardiman, Monica Olcina, Mark D. Robinson, Carste Krieg. Complement downregulation promotes an inflammatory signature that renders colorectal cancer susceptible to immunotherapy [abstract]. In: Proceedings of the AACR Special Conference on Colorectal Cancer; 2022 Oct 1-4; Portland, OR. Philadelphia (PA): AACR; Cancer Res 2022;82(23 Suppl_1):Abstract nr B006.

The accumulation of circulating low-density neutrophils (LDN) has been described in cancer patients and associated with tumor-supportive properties, as opposed to the high-density neutrophils (HDN). Here we aimed to evaluate the clinical significance of circulating LDN in lung cancer patients, and further assessed its diagnostic vs prognostic value. Using mass cytometry (CyTOF), we identified major subpopulations within the circulating LDN/HDN subsets and determined phenotypic modulations of these subsets along tumor progression. LDN were highly enriched in the low-density (LD) fraction of advanced lung cancer patients (median 7.0%; range 0.2%-80%, n = 64), but not in early stage patients (0.7%; 0.05%-6%; n = 35), healthy individuals (0.8%; 0%-3.5%; n = 15), or stable chronic obstructive pulmonary disease (COPD) patients (1.2%; 0.3%-7.4%, n = 13). Elevated LDN (>10%) remarkably related with poorer prognosis in late stage patients. We identified three main neutrophil subsets which proportions are markedly modified in cancer patients, with CD66b+ /CD10low /CXCR4+ /PDL1inter subset almost exclusively found in advanced lung cancer patients. We found substantial variability in subsets between patients, and demonstrated that HDN and LDN retain a degree of inherent spontaneous plasticity. Deep phenotypic characterization of cancer-related circulating neutrophils and their modulation along tumor progression is an important advancement in understanding the role of myeloid cells in lung cancer.

BackgroundSpatial transcriptomics is a next-generation sequencing technology that combines the strengths of transcriptome-wide RNA-sequencing with histological imaging to generate spatial maps of gene expression in intact tissue sections. The 10x Genomics Visium and Visium-Immunofluorescence (Visium-IF) platforms are widely available commercial technologies for quantifying spatially-resolved gene expression. These technologies directly couple gene expression with high resolution histological or immunofluorescence images that contain rich morphological information about the tissue section. However, extracting and integrating image features with gene expression data remains challenging.ResultsUsing MATLAB, we developed VistoSeg, which is a pipeline to process, analyze, and interactively visualize the high-resolution images from the 10x Genomics Visium and Visium-IF platforms. The output from VistoSeg can then be integrated with the spatial-molecular information in downstream analyses using common programming languages, such as R or Python.ConclusionVistoSeg provides user-friendly tools for integrating image-derived metrics from histological and immunofluorescent images with spatially-resolved gene expression data. This integrated approach can advance our understanding of the transcriptional landscape within tissue architecture. VistoSeg is freely available at http://research.libd.org/VistoSeg/.Impact StatementTechnologies for measuring gene activity levels, referred to as gene expression, have been evolving over decades and are the core of the transcriptomics subfield within genomics. The first report describing individual cell gene expression is from 2009 and as a method it became commercially available in 2014. While single cell transcriptomics increased our resolution beyond homogenate tissue, the advent of spatial transcriptomics technologies and commercial availability of spatial gene expression platforms, such as Visium, has facilitated studying gene expression in anatomical context. Visium measures local gene expression within the histological organization of single 6.5 mm2 cryosection of tissue. Spatially-resolved transcriptomics provides a new challenge: integrating spatial gene expression with high resolution tissue images (brightfield histology or fluorescent antibody staining). VistoSeg image processing software is compatible with both Visium and Visium-IF from 10x Genomics, which are spatially-resolved transcriptomics assays employing histological and immunofluorescent images, respectively. From these images, the number of cells, identity of cell types, and other image-derived markers can be obtained for thousands of 2,375 µm2 spots, where genome-wide gene expression is also measured. VistoSeg provides tools that enable processing these images in the context of gene expression maps to integrate these two high dimensional data types, and thus help unlock the new frontier in transcriptomics.

Several lines of evidence suggest that inflammation plays a pivotal role in the development and progression of colorectal cancer (CRC) and can be unleashed by the loss of innate immunosurveillance. The complement system is a well characterized first line of defense against pathogens and a central component of the immune response. Emerging evidence suggests that complement anaphylatoxin C3a produced upon complement activation and acting via its receptor (C3aR) may play a role in intestinal homeostasis. However, to date, it is unknown whether and how the C3a/C3aR axis can affect CRC. By mining publicly available datasets, we found that CpG island methylation of c3ar1 occurs in CRC patients and is associated with significant downregulation of C3aR. By reverse-translating this finding we were able to shift in APCMin/+ mice the tumorigenesis from the small intestine to the colon therefore generating a novel mouse model, which more closely mirrors the CRC in humans. Transcriptomic analysis on colorectal polyps from our newly developed genetic mouse model revealed a significant increase in innate and adaptive immune signatures in absence of C3aR. Furthermore, loss of C3aR significantly impacted the fecal and tumor-associated microbiota and supported the blooming of pro-inflammatory bacterial species as confirmed by experiments of fecal microbiota transplantation.Future studies will elucidate whether loss of C3aR can be exploited as a biomarker for sub-groups of CRC and whether the C3a/C3aR axis may be exploited for the generation of more effective therapeutic interventions.

We present distinct, a general method for differential analysis of full distributions that is well suited to applications on single-cell data, such as single-cell RNA sequencing and high-dimensional flow or mass cytometry data. High-throughput single-cell data reveal an unprecedented view of cell identity and allow complex variations between conditions to be discovered; nonetheless, most methods for differential expression target differences in the mean and struggle to identify changes where the mean is only marginally affected. distinct is based on a hierarchical non-parametric permutation approach and, by comparing empirical cumulative distribution functions, identifies both differential patterns involving changes in the mean, as well as more subtle variations that do not involve the mean. We performed extensive bench-marks across both simulated and experimental datasets from single-cell RNA sequencing and mass cytometry data, where distinct shows favourable performance, identifies more differential patterns than competitors, and displays good control of false positive and false discovery rates. distinct is available as a Bioconductor R package.

The role of inflammation in colorectal cancer (CRC) development and response to therapy is highly debated. While inflammation is a known driver of CRC, inflammatory immune infiltrates are a positive prognostic factor in CRC as inflammation can enhance response to immune checkpoint blockade (ICB) therapy. Yet, the majority of CRC are poorly infiltrated and even in presence of an inflammatory infiltrate half of these cancers become refractory to ICB. Therefore, there is a need to identify mechanisms to establish therapy supportive inflammation. We have previously reported that certain patients with CRC down-regulate the receptor for the complement anaphylatoxin C3a (C3aR). The complement system is a first line of defense against pathogens and a central component of the immune response. Emerging evidence suggests that C3aR may play a role in intestinal homeostasis. However, to date, it is unknown how C3aR affects CRC development and response to therapy with ICB. Using our novel mouse model of CRC (APCMin/+/C3aR−/−) we showed that loss of C3aR results in enhanced immune infiltration in typically “cold” tumors and changes in the intestinal microbiota. Similarly, we found that down-regulation of the complement C3aR in patients correlates with changes in specific tumor-infiltrating immune cells. Notably, when treated with a-PD1, APCMin/+/C3aR−/− mice showed a significant reduction of tumors, suggesting that the loss of C3aR unleashes in the colon a microbiota-mediated immune response that can be exploited for therapeutic purposes. Our findings reveal that C3aR may act as a previously unrecognized checkpoint to enhance antitumor immunity in CRC and that C3aR inhibition may be used to overcome ICB resistance in a larger group of CRC patients. This project was supported in part by NIH P20 GM120475 (core facility) at the MUSC Digestive Disease Research Core Center.

In the version of this article initially published, Figs. 5a,c and 6a were incorrect because of an error in a metadata spreadsheet that led to the healthy donor patient 2 (HD2) samples being used twice in the analysis of baseline samples and in the analysis at 12 weeks of anti-PD-1 therapy, while HD3 samples had not been used.

Benchmarking is a crucial step during computational analysis and method development. Recently, a number of new methods have been developed for analyzing high-dimensional cytometry data. However, it can be difficult for analysts and developers to find and access well-characterized benchmark datasets. Here, we present HDCytoData, a Bioconductor package providing streamlined access to several publicly available high-dimensional cytometry benchmark datasets. The package is designed to be extensible, allowing new datasets to be contributed by ourselves or other researchers in the future. Currently, the package includes a set of experimental and semi-simulated datasets, which have been used in our previous work to evaluate methods for clustering and differential analyses. Datasets are formatted into standard SummarizedExperiment and flowSet Bioconductor object formats, which include complete metadata within the objects. Access is provided through Bioconductor's ExperimentHub interface. The package is freely available from http://bioconductor.org/packages/HDCytoData.

Recent technological developments in high-dimensional flow cytometry and mass cytometry (CyTOF) have made it possible to detect expression levels of dozens of protein markers in thousands of cells per second, allowing cell populations to be characterized in unprecedented detail. Traditional data analysis by "manual gating" can be inefficient and unreliable in these high-dimensional settings, which has led to the development of a large number of automated analysis methods. Methods designed for unsupervised analysis use specialized clustering algorithms to detect and define cell populations for further downstream analysis. Here, we have performed an up-to-date, extensible performance comparison of clustering methods for high-dimensional flow and mass cytometry data. We evaluated methods using several publicly available data sets from experiments in immunology, containing both major and rare cell populations, with cell population identities from expert manual gating as the reference standard. Several methods performed well, including FlowSOM, X-shift, PhenoGraph, Rclusterpp, and flowMeans. Among these, FlowSOM had extremely fast runtimes, making this method well-suited for interactive, exploratory analysis of large, high-dimensional data sets on a standard laptop or desktop computer. These results extend previously published comparisons by focusing on high-dimensional data and including new methods developed for CyTOF data. R scripts to reproduce all analyses are available from GitHub (https://github.com/lmweber/cytometry-clustering-comparison), and pre-processed data files are available from FlowRepository (FR-FCM-ZZPH), allowing our comparisons to be extended to include new clustering methods and reference data sets. © 2016 The Authors. Cytometry Part A published by Wiley Periodicals, Inc. on behalf of ISAC.

Hartmann et al. show that, in narcolepsy, T cells exhibit a proinflammatory signature characterized by increased production of TNF, IL-2, and B cell–supporting cytokines., Narcolepsy type 1 is a devastating neurological sleep disorder resulting from the destruction of orexin-producing neurons in the central nervous system (CNS). Despite its striking association with the HLA-DQB1*06:02 allele, the autoimmune etiology of narcolepsy has remained largely hypothetical. Here, we compared peripheral mononucleated cells from narcolepsy patients with HLA-DQB1*06:02-matched healthy controls using high-dimensional mass cytometry in combination with algorithm-guided data analysis. Narcolepsy patients displayed multifaceted immune activation in CD4+ and CD8+ T cells dominated by elevated levels of B cell–supporting cytokines. Additionally, T cells from narcolepsy patients showed increased production of the proinflammatory cytokines IL-2 and TNF. Although it remains to be established whether these changes are primary to an autoimmune process in narcolepsy or secondary to orexin deficiency, these findings are indicative of inflammatory processes in the pathogenesis of this enigmatic disease.

High-dimensional flow and mass cytometry allow cell types and states to be characterized in great detail by measuring expression levels of more than 40 targeted protein markers per cell at the single-cell level. However, data analysis can be difficult, due to the large size and dimensionality of datasets as well as limitations of existing computational methods. Here, we present diffcyt, a new computational framework for differential discovery analyses in high-dimensional cytometry data, based on a combination of high-resolution clustering and empirical Bayes moderated tests adapted from transcriptomics. Our approach provides improved statistical performance, including for rare cell populations, along with flexible experimental designs and fast runtimes in an open-source framework., Lukas Weber et al. present diffcyt, an open-source computational framework for differential analysis of high-dimensional flow and mass cytometry data. The diffcyt framework uses high-resolution clustering and empirical Bayes moderated tests to provide improved statistical performance, flexible experimental designs, and fast runtimes.

CRISPR-Cas9 enables efficient sequence-specific mutagenesis for creating somatic or germline mutants of model organisms. Key constraints in vivo remain the expression and delivery of active Cas9-guideRNA ribonucleoprotein complexes (RNPs) with minimal toxicity, variable mutagenesis efficiencies depending on targeting sequence, and high mutation mosaicism. Here, we apply in vitro-assembled, fluorescent Cas9-sgRNA RNPs in solubilizing salt solution to achieve maximal mutagenesis efficiency in zebrafish embryos. MiSeq-based sequence analysis of targeted loci in individual embryos using CrispRVariants, a customized software tool for mutagenesis quantification and visualization, reveals efficient bi-allelic mutagenesis that reaches saturation at several tested gene loci. Such virtually complete mutagenesis exposes loss-of-function phenotypes for candidate genes in somatic mutant embryos for subsequent generation of stable germline mutants. We further show that targeting of non-coding elements in gene-regulatory regions using saturating mutagenesis uncovers functional control elements in transgenic reporters and endogenous genes in injected embryos. Our results establish that optimally solubilized, in vitro assembled fluorescent Cas9-sgRNA RNPs provide a reproducible reagent for direct and scalable loss-of-function studies and applications beyond zebrafish experiments that require maximal DNA cutting efficiency in vivo.

Immune-checkpoint blockade has revolutionized cancer therapy. In particular, inhibition of programmed cell death protein 1 (PD-1) has been found to be effective for the treatment of metastatic melanoma and other cancers. Despite a dramatic increase in progression-free survival, a large proportion of patients do not show durable responses. Therefore, predictive biomarkers of a clinical response are urgently needed. Here we used high-dimensional single-cell mass cytometry and a bioinformatics pipeline for the in-depth characterization of the immune cell subsets in the peripheral blood of patients with stage IV melanoma before and after 12 weeks of anti-PD-1 immunotherapy. During therapy, we observed a clear response to immunotherapy in the T cell compartment. However, before commencing therapy, a strong predictor of progression-free and overall survival in response to anti-PD-1 immunotherapy was the frequency of CD14+CD16-HLA-DRhi monocytes. We confirmed this by conventional flow cytometry in an independent, blinded validation cohort, and we propose that the frequency of monocytes in PBMCs may serve in clinical decision support.

Periodontal disease of equine cheek teeth is common and may lead to tooth loss if left untreated. Limited information is available comparing the effectiveness of treatment methods. The objective of this study was to retrospectively compare the effectiveness of 4 commonly used treatments in reducing periodontal pocket depth (in addition to routine dental treatment and occlusal equilibration). The 4 treatments compared were (1) removal of feed material, lavaging the pocket with dilute chlorhexidine, and then rinsing the mouth with chlorhexidine-containing mouthwash (CL); (2) CL plus placement of metronidazole into the pocket (M); (3) M plus the addition of polyvinyl siloxane temporary filling over the diastema (PVS); and (4) diastema widening to increase the interdental space, then PVS (DW). Pocket measurements were compared before and 2 to 6 months after treatment. Treatment groups CL, M, and PVS showed statistically significant reductions in pocket depth following treatment. The mean pocket depth reduction was the greatest in the DW group (and this was the only group with no cases having an increase in pocket depth), but this was not significant due to the small sample size. Additional analysis to compare effectiveness revealed a confounding effect of initial pocket depth. After accounting for this, DW was associated with smaller improvements than the other treatments, however, this was also based on a small sample size. After accounting for confounders, differences between treatments CL, M and PVS were not found to be significant, although all were associated with statistically significant reductions in pocket depth.

High-dimensional mass and flow cytometry (HDCyto) experiments have become a method of choice for high-throughput interrogation and characterization of cell populations. Here, we present an updated R-based pipeline for differential analyses of HDCyto data, largely based on Bioconductor packages. We computationally define cell populations using FlowSOM clustering, and facilitate an optional but reproducible strategy for manual merging of algorithm-generated clusters. Our workflow offers different analysis paths, including association of cell type abundance with a phenotype or changes in signalling markers within specific subpopulations, or differential analyses of aggregated signals. Importantly, the differential analyses we show are based on regression frameworks where the HDCyto data is the response; thus, we are able to model arbitrary experimental designs, such as those with batch effects, paired designs and so on. In particular, we apply generalized linear mixed models or linear mixed models to analyses of cell population abundance or cell-population-specific analyses of signaling markers, allowing overdispersion in cell count or aggregated signals across samples to be appropriately modeled. To support the formal statistical analyses, we encourage exploratory data analysis at every step, including quality control (e.g., multi-dimensional scaling plots), reporting of clustering results (dimensionality reduction, heatmaps with dendrograms) and differential analyses (e.g., plots of aggregated signals).

Recent technological developments in high-dimensional flow cytometry and mass cytometry (CyTOF) have made it possible to detect expression levels of dozens of protein markers in thousands of cells per second, allowing cell populations to be characterized in unprecedented detail. Traditional data analysis by "manual gating" can be inefficient and unreliable in these high-dimensional settings, which has led to the development of a large number of automated analysis methods. Methods designed for unsupervised analysis use specialized clustering algorithms to detect and define cell populations for further downstream analysis. Here, we have performed an up-to-date, extensible performance comparison of clustering methods for high-dimensional flow and mass cytometry data. We evaluated methods using several publicly available data sets from experiments in immunology, containing both major and rare cell populations, with cell population identities from expert manual gating as the reference standard. Several methods performed well, including FlowSOM, X-shift, PhenoGraph, Rclusterpp, and flowMeans. Among these, FlowSOM had extremely fast runtimes, making this method well-suited for interactive, exploratory analysis of large, high-dimensional data sets on a standard laptop or desktop computer. These results extend previously published comparisons by focusing on high-dimensional data and including new methods developed for CyTOF data. R scripts to reproduce all analyses are available from GitHub (https://github.com/lmweber/cytometry-clustering-comparison), and pre-processed data files are available from FlowRepository (FR-FCM-ZZPH), allowing our comparisons to be extended to include new clustering methods and reference data sets. © 2016 The Authors. Cytometry Part A published by Wiley Periodicals, Inc. on behalf of ISAC.

Checkpoint inhibition of programmed cell death protein 1 (PD-1) has revolutionized cancer therapy. As more and more patients with metastatic melanoma but also other cancers show a dramatic increase in progression-free survival, still a large proportion of patients do not show durable response. As options for monotherapies, as well as combination immunotherapies, increase in the clinics, predictive biomarkers of clinical response are urgently needed. We employed high-dimensional single cell mass cytometry and a machine-learning bioinformatics pipeline for the in-depth characterization of all immune cells in the peripheral blood of stage IV melanoma patients before and at 12 weeks of anti-PD-1 immunotherapy. We observed a clear treatment response of patients under immunotherapy. However, a strong predictor of progression free and overall survival in response to anti-PD-1 immunotherapy before therapy was the frequency of CD14+CD16-HLA-DRhi monocytes. We could confirm this by conventional flow cytometry in an independent, blinded validation cohort and propose this as a novel predictive biomarker for therapy decisions in the clinic. We further present a high dimensional analysis workflow to analyze predictive and prognostic immune biomarkers during future clinical trials. Citation Format: Carsten Krieg, Malgorzata Nowicka, Silvia Guglietta, Sabrina Schindler, Felix J. Hartmann, Lukas M. Weber, Reinhard Dummer, Mark D. Robinson, Mitchell P. Levesque, Burkhard Becher. High dimensional single cell analysis predicts response to anti-PD-1 immunotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 566.

Blockade of checkpoint inhibitors such as programmed cell death protein 1 (PD-1) set the pace for the immunotherapy of cancer. Following the initial success in metastatic melanoma, by showing a dramatic increase in progression-free survival, PD-1 immunotherapy was expanded to other cancers. As PD-1 monotherapy advances into first line treatment and combination of checkpoint inhibitors and other therapies start to show even increased response rates, still a large proportion of patients do not show durable responses. Therefore, predictive biomarkers of clinical response for stratification of patients most likely to respond to immunotherapy are urgently needed. In this study, we employed high-dimensional single cell mass cytometry and a generalized machine-learning bioinformatics workflow for the in-depth characterization of all immune cells in the peripheral blood of stage IV melanoma patients before and at 12 weeks of anti-PD-1 immunotherapy. In agreement with published literature a clear response in the T cell compartment of patients under immunotherapy was observed. To our surprise, prior to commencing PD-1 immunotherapy, a strong predictor of progression free and overall survival in responders to anti-PD-1 was a higher frequency of CD14+CD16−HLA-DRhi monocytes. We confirmed this finding by conventional flow cytometry in an independent, blinded validation cohort. Altogether, in this study, we identified for the first time a novel predictive biomarker for therapy decision in the treatment of metastatic melanoma and propose to extend this approach to select for positive responders in other cancers prior to immune checkpoint therapy.

Reasons for performing the study. Equine cheek teeth periodontal disease is a common, often painful dental condition that may lead to premature tooth loss if left untreated. All cases of periodontal disease in this study were associated with diastemata, with the majority of cases being secondary or senile diastemata. At present limited objective information is available comparing different treatment methods for the condition to assist clinicians in making evidence-based treatment decisions. Objectives. To compare the effectiveness of four commonly used treatments for equine periodontal disease (additional to routine dental treatment and equilibration) to assess their merit in reducing periodontal pocket depth. Study design. Retrospective case series. Methods. Four commonly used equine periodontal treatments were compared: Removing feed material and lavaging the periodontal pocket with dilute chlorhexidine and rinsing the mouth with chlorhexidine based mouthwash (‘Hexarinse’), ‘clean and antiseptic lavage’ (CL); CL plus the addition of metronidazole antibiotics into the periodontal pockets (M); M plus the addition of a polyvinyl siloxane temporary filling over the diastema (PVS); and diastema widening to open the diastema and increase the interdental space between adjacent teeth, then PVS (DW). Periodontal pocket depth measurements were compared before and after treatment at treatment intervals between two and six months. Results. Treatment groups CL, M and PVS showed statistically significant reductions in pocket depth following treatment. Mean pocket depth reduction was greatest in the DW group (and this was the only group with no cases involving an increase in pocket depth), but this was based on a small sample size and was not statistically significant. Additional analysis to compare effectiveness revealed a strong confounding effect of initial pocket depth. After taking this into account, there was some evidence that DW was associated with smaller improvements than the other treatments, however this result was also based on a small sample size, and influenced by a small number of cases with particularly large improvements. Among the other treatments, no statistically significant differences in effectiveness were found. Conclusion. This study has shown that treatment methods CL, M and PVS are associated with statistically significant reductions in pocket depth. Due to the confounding effect of initial pocket depth, no clear differences in effectiveness were found between treatment methods.

High dimensional mass and flow cytometry (HDCyto) experiments have become a method of choice for high throughput interrogation and characterization of cell populations.Here, we present an R-based pipeline for differential analyses of HDCyto data, largely based on Bioconductor packages. We computationally define cell populations using FlowSOM clustering, and facilitate an optional but reproducible strategy for manual merging of algorithm-generated clusters. Our workflow offers different analysis paths, including association of cell type abundance with a phenotype or changes in signaling markers within specific subpopulations, or differential analyses of aggregated signals. Importantly, the differential analyses we show are based on regression frameworks where the HDCyto data is the response; thus, we are able to model arbitrary experimental designs, such as those with batch effects, paired designs and so on. In particular, we apply generalized linear mixed models to analyses of cell population abundance or cell-population-specific analyses of signaling markers, allowing overdispersion in cell count or aggregated signals across samples to be appropriately modeled. To support the formal statistical analyses, we encourage exploratory data analysis at every step, including quality control (e.g. multi-dimensional scaling plots), reporting of clustering results (dimensionality reduction, heatmaps with dendrograms) and differential analyses (e.g. plots of aggregated signals).

Checkpoint inhibitor therapy (CIT) has revolutionized cancer therapy. Notably, inhibition of programmed cell death protein 1 (PD-1) showed benefit in the treatment of metastatic melanoma and has even been promoted to first line therapy in advanced melanoma. Despite an increase in overall survival a minority of patients profits from CIT with anti-PD-1. Therefore, immune components or markers predicting responsiveness are urgently needed. Using phenotypic and functional markers we used high-dimensional single cell cytometry by time of flight (CyTOF), to measure blood samples from anti-PD-1 treated patients before and during therapy. CyTOF uses metal-tagged instead of fluorochrome-tagged antibodies, thus eliminating spectral overlap and, in combination with barcoding, enables the measurement of more than 30 parameters on a single cell in multiple samples in parallel. Due to the complexity of the data set and in order to reduce analytical bias we adapted a custom bioinformatic workflow. Using this approach we find immune signatures which allow us to separate responders from non-responders and might provide insights into the biology of immune therapy.

Reasons for performing the study. Equine periodontal disease is a common, often painful dental condition that may lead to premature tooth loss if left untreated [1,2,3]. At present little research has been done comparing different treatment methods for the condition to assist clinicians in making educated treatment decisions. Objectives. To compare the effectiveness of four commonly used treatment options for periodontal disease (on top of routine dental treatment and equilibration) to assess their merit in reducing periodontal pocket depth. Study Design. Retrospective case series. Methods. Four commonly used treatment options were compared: Removing feed material and disinfecting the periodontal pocket with dilute chlorhexidine and rinsing the mouth with chlorhexidine based mouthwash (‘Hexarinse’), ‘cleaned and disinfected’ (CD); CD plus the addition of metronidazole antibiotics into the periodontal pockets (M); M plus the addition of a polyvinyl siloxane temporary filling over the diastema (PVS); and diastema burring to widen the gap between the teeth, then PVS (DB). Pocket depth measurements were compared before and after treatment at treatment intervals between two and six months. Results. Treatment groups CD, M and PVS showed statistically significant reductions in pocket depth between visits. Mean reduction was greatest in the DB group, but this was not statistically significant. Additional analysis to compare effectiveness revealed a strong confounding effect of initial pocket depth. After taking this into account, there was some evidence that DB was associated with smaller improvements than the other treatments, however this result was based on a small sample size. Among the other treatments, no statistically significant differences in effectiveness were found. Conclusion. This study has shown that treatment methods CD, M and PVS are associated with statistically significant reductions in pocket depth. Due to the confounding effect of initial pocket depth, no clear differences in effectiveness were found between treatment methods.

DNA methylation, the reversible addition of methyl groups at CpG dinucleotides, represents an important regulatory layer associated with gene expression. Changed methylation status has been noted across diverse pathological states, including cancer. The rapid development and uptake of microarrays and large scale DNA sequencing has prompted an explosion of data analytic methods for processing and discovering changes in DNA methylation across varied data types. In this mini-review, we present a compact and accessible discussion of many of the salient challenges, such as experimental design, statistical methods for differential methylation detection, critical considerations such as cell type composition and the potential confounding that can arise from batch effects. From a statistical perspective, our main interests include the use of empirical Bayes or hierarchical models, which have proved immensely powerful in genomics, and the procedures by which false discovery control is achieved.

In the political economy model of Grossman and Helpman (1995), two incumbent governments attempt to negotiate a free trade agreement (FTA), while special interest groups in each country influence negotiations by offering financial contributions to their governments. As a consequence, a set of politically sensitive industries is excluded from the proposed FTA. Using the empirical methodology of Gawande, Sanguinetti, and Bohara (2001), this paper shows that the Grossman-Helpman (1995) model successfully predicts the set of excluded industries for the recently implemented Australia-United States FTA. It is also shown that the set of exclusions favours Australian interest groups, which could indicate that the gains from the FTA are lower for the government of Australia than for the government of the United States.

DNA methylation, and specifically the reversible addition of methyl groups at CpG dinucleotides genome-wide, represents an important layer that is associated with the regulation of gene expression. In particular, aberrations in the methylation status have been noted across a diverse set of pathological states, including cancer. With the rapid development and uptake of large scale sequencing of short DNA fragments, there has been an explosion of data analytic methods for processing and discovering changes in DNA methylation across diverse data types. In this mini-review, we aim to condense many of the salient challenges, such as experimental design, statistical methods for differential methylation detection and critical considerations such as cell type composition and the potential confounding that can arise from batch effects, into a compact and accessible format. Our main interests, from a statistical perspective, include the practical use of empirical Bayes or hierarchical models, which have been shown to be immensely powerful and flexible in genomics and the procedures by which control of false discoveries are made. Of course, there are many critical platform-specific data preprocessing aspects that we do not discuss here. In addition, we do not make formal performance comparisons of the methods, but rather describe the commonly used statistical models and many of the pertinent issues; we make some recommendations for further study.