Optimal adjustment sets for causal query estimation in partially observed biomolecular networks.

Causal query estimation in biomolecular networks commonly selects a 'valid adjustment set', i.e. a subset of network variables that eliminates the bias of the estimator. A same query may have multiple valid adjustment sets, each with a different variance. When networks are partially observed, current methods use graph-based criteria to find an adjustment set that minimizes asymptotic variance. Unfortunately, many models that share the same graph topology, and therefore same functional dependencies, may differ in the processes that generate the observational data. In these cases, the topology-based criteria fail to distinguish the variances of the adjustment sets. This deficiency can lead to sub-optimal adjustment sets, and to miss-characterization of the effect of the intervention. We propose an approach for deriving 'optimal adjustment sets' that takes into account the nature of the data, bias and finite-sample variance of the estimator, and cost. It empirically learns the data generating processes from historical experimental data, and characterizes the properties of the estimators by simulation. We demonstrate the utility of the proposed approach in four biomolecular Case studies with different topologies and different data generation processes. The implementation and reproducible Case studies are at https://github.com/srtaheri/OptimalAdjustmentSet.

CD69 marks a subpopulation of acute myeloid leukemia with enhanced colony forming capacity and a unique signaling activation state.

In acute myeloid leukemia (AML), leukemia stem cells (LSCs) have self-renewal potential and are responsible for relapse. We previously showed that, in Mll-AF9/NRASG12V murine AML, CD69 expression marks an LSC-enriched subpopulation with enhanced in vivo self-renewal capacity. Here, we used CyTOF to define activated signaling pathways in LSC subpopulations in Mll-AF9/NRASG12V AML. Furthermore, we compared the signaling activation states of CD69High and CD36High subsets of primary human AML. The human CD69High subset expresses low levels of Ki67 and high levels of NFκB and pMAPKAPKII. Additionally, the human CD69High AML subset also has enhanced colony-forming capacity. We applied Bayesian network modeling to compare the global signaling network within the human AML subsets. We find that distinct signaling states, distinguished by NFκB and pMAPKAPKII levels, correlate with divergent functional subsets, defined by CD69 and CD36 expression, in human AML. Targeting NFκB with proteasome inhibition diminished colony formation.

Immunophenotypically-defined murine AML stem cells harbor self-renewing and non-self-renewing subsets that display unique signaling characteristics.CD69, an NFκB target gene, marks a subset of human AML with increased colony forming capacity and reduced proliferation.NFκB activation correlates with the global signaling pathway activation state in human AML.

Peripheral T-Cells, B-Cells, and Monocytes from Multiple Sclerosis Patients Supplemented with High-Dose Vitamin D Show Distinct Changes in Gene Expression Profiles.

Vitamin D is a steroid hormone that has been widely studied as a potential therapy for multiple sclerosis and other inflammatory disorders. Pre-clinical studies have implicated vitamin D in the transcription of thousands of genes, but its influence may vary by cell type. A handful of clinical studies have failed to identify an in vivo gene expression signature when using bulk analysis of all peripheral immune cells. We hypothesized that vitamin D's gene signature would vary by immune cell type, requiring the analysis of distinct cell types. Multiple sclerosis patients (n = 18) were given high-dose vitamin D (10,400 IU/day) for six months as part of a prospective clinical trial (NCT01024777). We collected peripheral blood mononuclear cells from participants at baseline and again after six months of treatment. We used flow cytometry to isolate three immune cell types (CD4+ T-cells, CD19+ B-cells, CD14+ monocytes) for RNA microarray analysis and compared the expression profiles between baseline and six months. We identified distinct sets of differentially expressed genes and enriched pathways between baseline and six months for each cell type. Vitamin D's in vivo gene expression profile in the immune system likely differs by cell type. Future clinical studies should consider techniques that allow for a similar cell-type resolution.

Do-calculus enables estimation of causal effects in partially observed biomolecular pathways.

MOTIVATION: Estimating causal queries, such as changes in protein abundance in response to a perturbation, is a fundamental task in the analysis of biomolecular pathways. The estimation requires experimental measurements on the pathway components. However, in practice many pathway components are left unobserved (latent) because they are either unknown, or difficult to measure. Latent variable models (LVMs) are well-suited for such estimation. Unfortunately, LVM-based estimation of causal queries can be inaccurate when parameters of the latent variables are not uniquely identified, or when the number of latent variables is misspecified. This has limited the use of LVMs for causal inference in biomolecular pathways. RESULTS: In this article, we propose a general and practical approach for LVM-based estimation of causal queries. We prove that, despite the challenges above, LVM-based estimators of causal queries are accurate if the queries are identifiable according to Pearl's do-calculus and describe an algorithm for its estimation. We illustrate the breadth and the practical utility of this approach for estimating causal queries in four synthetic and two experimental case studies, where structures of biomolecular pathways challenge the existing methods for causal query estimation. AVAILABILITY AND IMPLEMENTATION: The code and the data documenting all the case studies are available at https://github.com/srtaheri/LVMwithDoCalculus. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Variation of Immune Cell Responses in Humans Reveals Sex-Specific Coordinated Signaling Across Cell Types.

Assessing the health and competence of the immune system is central to evaluating vaccination responses, autoimmune conditions, cancer prognosis, and treatment. With an increasing number of studies examining immune dysregulation, there is a growing need for a curated reference of variation in immune parameters in healthy individuals. We used mass cytometry (CyTOF) to profile blood from 86 humans in response to 15 ex vivo immune stimuli. We present reference ranges for cell-specific immune markers and highlight differences that appear across sex and age. We identified modules of immune features that suggest there exists an underlying structure to the immune system based on signaling pathway responses across cell types. We observed increased MAPK signaling in inflammatory pathways in innate immune cells and greater overall coordination of immune cell responses in females. In contrast, males exhibited stronger pSTAT1 and pTBK1 responses. These reference data are publicly available as a resource for immune profiling studies.

Identifying patterns in amyotrophic lateral sclerosis progression from sparse longitudinal data.

The clinical presentation of amyotrophic lateral sclerosis (ALS), a fatal neurodegenerative disease, varies widely across patients, making it challenging to determine if potential therapeutics slow progression. We sought to determine whether there were common patterns of disease progression that could aid in the design and analysis of clinical trials. We developed an approach based on a mixture of Gaussian processes to identify clusters of patients sharing similar disease progression patterns, modeling their average trajectories and the variability in each cluster. We show that ALS progression is frequently nonlinear, with periods of stable disease preceded or followed by rapid decline. We also show that our approach can be extended to Alzheimer's and Parkinson's diseases. Our results advance the characterization of disease progression of ALS and provide a flexible modeling approach that can be applied to other progressive diseases.

An integrated multi-omic analysis of iPSC-derived motor neurons from C9ORF72 ALS patients.

Neurodegenerative diseases are challenging for systems biology because of the lack of reliable animal models or patient samples at early disease stages. Induced pluripotent stem cells (iPSCs) could address these challenges. We investigated DNA, RNA, epigenetics, and proteins in iPSC-derived motor neurons from patients with ALS carrying hexanucleotide expansions in C9ORF72. Using integrative computational methods combining all omics datasets, we identified novel and known dysregulated pathways. We used a C9ORF72 Drosophila model to distinguish pathways contributing to disease phenotypes from compensatory ones and confirmed alterations in some pathways in postmortem spinal cord tissue of patients with ALS. A different differentiation protocol was used to derive a separate set of C9ORF72 and control motor neurons. Many individual -omics differed by protocol, but some core dysregulated pathways were consistent. This strategy of analyzing patient-specific neurons provides disease-related outcomes with small numbers of heterogeneous lines and reduces variation from single-omics to elucidate network-based signatures.

High-grade serous ovarian tumor cells modulate NK cell function to create an immune-tolerant microenvironment.

Tubo-ovarian high-grade serous carcinoma (HGSC) is unresponsive to immune checkpoint blockade despite significant frequencies of exhausted T cells. Here we apply mass cytometry and uncover decidual-like natural killer (dl-NK) cell subpopulations (CD56+CD9+CXCR3+KIR+CD3-CD16-) in newly diagnosed HGSC samples that correlate with both tumor and transitioning epithelial-mesenchymal cell abundance. We show different combinatorial expression patterns of ligands for activating and inhibitory NK receptors within three HGSC tumor compartments: epithelial (E), transitioning epithelial-mesenchymal (EV), and mesenchymal (vimentin expressing [V]), with a more inhibitory ligand phenotype in V cells. In cocultures, NK-92 natural killer cells acquire CD9 from HGSC tumor cells by trogocytosis, resulting in reduced anti-tumor cytokine production and cytotoxicity. Cytotoxicity in these cocultures is restored with a CD9-blocking antibody or CD9 CRISPR knockout, thereby identifying mechanisms of immune suppression in HGSC. CD9 is widely expressed in HGSC tumors and so represents an important new therapeutic target with immediate relevance for NK immunotherapy.

Single-Cell Gene Expression Analyses Reveal Distinct Self-Renewing and Proliferating Subsets in the Leukemia Stem Cell Compartment in Acute Myeloid Leukemia.

Standard chemotherapy for acute myeloid leukemia (AML) targets proliferative cells and efficiently induces complete remission; however, many patients relapse and die of their disease. Relapse is caused by leukemia stem cells (LSC), the cells with self-renewal capacity. Self-renewal and proliferation are separate functions in normal hematopoietic stem cells (HSC) in steady-state conditions. If these functions are also separate functions in LSCs, then antiproliferative therapies may fail to target self-renewal, allowing for relapse. We investigated whether proliferation and self-renewal are separate functions in LSCs as they often are in HSCs. Distinct transcriptional profiles within LSCs of Mll-AF9/NRASG12V murine AML were identified using single-cell RNA sequencing. Single-cell qPCR revealed that these genes were also differentially expressed in primary human LSCs and normal human HSPCs. A smaller subset of these genes was upregulated in LSCs relative to HSPCs; this subset of genes constitutes "LSC-specific" genes in human AML. To assess the differences between these profiles, we identified cell surface markers, CD69 and CD36, whose genes were differentially expressed between these profiles. In vivo mouse reconstitution assays resealed that only CD69High LSCs were capable of self-renewal and were poorly proliferative. In contrast, CD36High LSCs were unable to transplant leukemia but were highly proliferative. These data demonstrate that the transcriptional foundations of self-renewal and proliferation are distinct in LSCs as they often are in normal stem cells and suggest that therapeutic strategies that target self-renewal, in addition to proliferation, are critical to prevent relapse and improve survival in AML. SIGNIFICANCE: These findings define and functionally validate a self-renewal gene profile of leukemia stem cells at the single-cell level and demonstrate that self-renewal and proliferation are distinct in AML. GRAPHICAL ABSTRACT: http://cancerres.aacrjournals.org/content/canres/80/3/458/F1.large.jpg.

GateFinder: projection-based gating strategy optimization for flow and mass cytometry.

Motivation: High-parameter single-cell technologies can reveal novel cell populations of interest, but studying or validating these populations using lower-parameter methods remains challenging. Results: Here, we present GateFinder, an algorithm that enriches high-dimensional cell types with simple, stepwise polygon gates requiring only two markers at a time. A series of case studies of complex cell types illustrates how simplified enrichment strategies can enable more efficient assays, reveal novel biomarkers and clarify underlying biology. Availability and implementation: The GateFinder algorithm is implemented as a free and open-source package for BioConductor: https://nalab.stanford.edu/gatefinder. Supplementary information: Supplementary data are available at Bioinformatics online.

A Bayesian Active Learning Experimental Design for Inferring Signaling Networks.

Machine learning methods for learning network structure are applied to quantitative proteomics experiments and reverse-engineer intracellular signal transduction networks. They provide insight into the rewiring of signaling within the context of a disease or a phenotype. To learn the causal patterns of influence between proteins in the network, the methods require experiments that include targeted interventions that fix the activity of specific proteins. However, the interventions are costly and add experimental complexity. We describe an active learning strategy for selecting optimal interventions. Our approach takes as inputs pathway databases and historic data sets, expresses them in form of prior probability distributions on network structures, and selects interventions that maximize their expected contribution to structure learning. Evaluations on simulated and real data show that the strategy reduces the detection error of validated edges as compared with an unguided choice of interventions and avoids redundant interventions, thereby increasing the effectiveness of the experiment.

The Library of Integrated Network-Based Cellular Signatures NIH Program: System-Level Cataloging of Human Cells Response to Perturbations.

The Library of Integrated Network-Based Cellular Signatures (LINCS) is an NIH Common Fund program that catalogs how human cells globally respond to chemical, genetic, and disease perturbations. Resources generated by LINCS include experimental and computational methods, visualization tools, molecular and imaging data, and signatures. By assembling an integrated picture of the range of responses of human cells exposed to many perturbations, the LINCS program aims to better understand human disease and to advance the development of new therapies. Perturbations under study include drugs, genetic perturbations, tissue micro-environments, antibodies, and disease-causing mutations. Responses to perturbations are measured by transcript profiling, mass spectrometry, cell imaging, and biochemical methods, among other assays. The LINCS program focuses on cellular physiology shared among tissues and cell types relevant to an array of diseases, including cancer, heart disease, and neurodegenerative disorders. This Perspective describes LINCS technologies, datasets, tools, and approaches to data accessibility and reusability.

Distinct signaling programs control human hematopoietic stem cell survival and proliferation.

Several growth factors (GFs) that together promote quiescent human hematopoietic stem cell (HSC) expansion ex vivo have been identified; however, the molecular mechanisms by which these GFs regulate the survival, proliferation. and differentiation of human HSCs remain poorly understood. We now describe experiments in which we used mass cytometry to simultaneously measure multiple surface markers, transcription factors, active signaling intermediates, viability, and cell-cycle indicators in single CD34+ cord blood cells before and up to 2 hours after their stimulation with stem cell factor, Fms-like tyrosine kinase 3 ligand, interleukin-3, interleukin-6, and granulocyte colony-stimulating factor (5 GFs) either alone or combined. Cells with a CD34+CD38-CD45RA-CD90+CD49f+ (CD49f+) phenotype (∼10% HSCs with >6-month repopulating activity in immunodeficient mice) displayed rapid increases in activated STAT1/3/5, extracellular signal-regulated kinase 1/2, AKT, CREB, and S6 by 1 or more of these GFs, and ß-catenin only when the 5 GFs were combined. Certain minority subsets within the CD49f+ compartment were poorly GF-responsive and, among the more GF-responsive subsets of CD49f+ cells, different signaling intermediates correlated with the levels of the myeloid- and lymphoid-associated transcription factors measured. Phenotypically similar, but CD90-CD49f- cells (MPPs) contained lower baseline levels of multiple signaling intermediates than the CD90+CD49f+ cells, but showed similar response amplitudes to the same GFs. Importantly, we found activation or inhibition of AKT and ß-catenin directly altered immediate CD49f+ cell survival and proliferation. These findings identify rapid signaling events that 5 GFs elicit directly in the most primitive human hematopoietic cell types to promote their survival and proliferation.

From Correlation to Causality: Statistical Approaches to Learning Regulatory Relationships in Large-Scale Biomolecular Investigations.

Causal inference, the task of uncovering regulatory relationships between components of biomolecular pathways and networks, is a primary goal of many high-throughput investigations. Statistical associations between observed protein concentrations can suggest an enticing number of hypotheses regarding the underlying causal interactions, but when do such associations reflect the underlying causal biomolecular mechanisms? The goal of this perspective is to provide suggestions for causal inference in large-scale experiments, which utilize high-throughput technologies such as mass-spectrometry-based proteomics. We describe in nontechnical terms the pitfalls of inference in large data sets and suggest methods to overcome these pitfalls and reliably find regulatory associations.

NRASG12V oncogene facilitates self-renewal in a murine model of acute myelogenous leukemia.

Mutant RAS oncoproteins activate signaling molecules that drive oncogenesis in multiple human tumors including acute myelogenous leukemia (AML). However, the specific functions of these pathways in AML are unclear, thwarting the rational application of targeted therapeutics. To elucidate the downstream functions of activated NRAS in AML, we used a murine model that harbors Mll-AF9 and a tetracycline-repressible, activated NRAS (NRAS(G12V)). Using computational approaches to explore our gene-expression data sets, we found that NRAS(G12V) enforced the leukemia self-renewal gene-expression signature and was required to maintain an MLL-AF9- and Myb-dependent leukemia self-renewal gene-expression program. NRAS(G12V) was required for leukemia self-renewal independent of its effects on growth and survival. Analysis of the gene-expression patterns of leukemic subpopulations revealed that the NRAS(G12V)-mediated leukemia self-renewal signature is preferentially expressed in the leukemia stem cell-enriched subpopulation. In a multiplexed analysis of RAS-dependent signaling, Mac-1(Low) cells, which harbor leukemia stem cells, were preferentially sensitive to NRAS(G12V) withdrawal. NRAS(G12V) maintained leukemia self-renewal through mTOR and MEK pathway activation, implicating these pathways as potential targets for cancer stem cell-specific therapies. Together, these experimental results define a RAS oncogene-driven function that is critical for leukemia maintenance and represents a novel mechanism of oncogene addiction.

Normalization of mass cytometry data with bead standards.

Mass cytometry uses atomic mass spectrometry combined with isotopically pure reporter elements to currently measure as many as 40 parameters per single cell. As with any quantitative technology, there is a fundamental need for quality assurance and normalization protocols. In the case of mass cytometry, the signal variation over time due to changes in instrument performance combined with intervals between scheduled maintenance must be accounted for and then normalized. Here, samples were mixed with polystyrene beads embedded with metal lanthanides, allowing monitoring of mass cytometry instrument performance over multiple days of data acquisition. The protocol described here includes simultaneous measurements of beads and cells on the mass cytometer, subsequent extraction of the bead-based signature, and the application of an algorithm enabling correction of both short- and long-term signal fluctuations. The variation in the intensity of the beads that remains after normalization may also be used to determine data quality. Application of the algorithm to a one-month longitudinal analysis of a human peripheral blood sample reduced the range of median signal fluctuation from 4.9-fold to 1.3-fold.

Multiplexed mass cytometry profiling of cellular states perturbed by small-molecule regulators.

Mass cytometry facilitates high-dimensional, quantitative analysis of the effects of bioactive molecules on human samples at single-cell resolution, but instruments process only one sample at a time. Here we describe mass-tag cellular barcoding (MCB), which increases mass cytometry throughput by using n metal ion tags to multiplex up to 2n samples. We used seven tags to multiplex an entire 96-well plate, and applied MCB to characterize human peripheral blood mononuclear cell (PBMC) signaling dynamics and cell-to-cell communication, signaling variability between PBMCs from eight human donors, and the effects of 27 inhibitors on this system. For each inhibitor, we measured 14 phosphorylation sites in 14 PBMC types at 96 conditions, resulting in 18,816 quantified phosphorylation levels from each multiplexed sample. This high-dimensional, systems-level inquiry allowed analysis across cell-type and signaling space, reclassified inhibitors and revealed off-target effects. High-content, high-throughput screening with MCB should be useful for drug discovery, preclinical testing and mechanistic investigation of human disease.

Extracting a cellular hierarchy from high-dimensional cytometry data with SPADE.

The ability to analyze multiple single-cell parameters is critical for understanding cellular heterogeneity. Despite recent advances in measurement technology, methods for analyzing high-dimensional single-cell data are often subjective, labor intensive and require prior knowledge of the biological system. To objectively uncover cellular heterogeneity from single-cell measurements, we present a versatile computational approach, spanning-tree progression analysis of density-normalized events (SPADE). We applied SPADE to flow cytometry data of mouse bone marrow and to mass cytometry data of human bone marrow. In both cases, SPADE organized cells in a hierarchy of related phenotypes that partially recapitulated well-described patterns of hematopoiesis. We demonstrate that SPADE is robust to measurement noise and to the choice of cellular markers. SPADE facilitates the analysis of cellular heterogeneity, the identification of cell types and comparison of functional markers in response to perturbations.

Single-cell mass cytometry of differential immune and drug responses across a human hematopoietic continuum.

Flow cytometry is an essential tool for dissecting the functional complexity of hematopoiesis. We used single-cell "mass cytometry" to examine healthy human bone marrow, measuring 34 parameters simultaneously in single cells (binding of 31 antibodies, viability, DNA content, and relative cell size). The signaling behavior of cell subsets spanning a defined hematopoietic hierarchy was monitored with 18 simultaneous markers of functional signaling states perturbed by a set of ex vivo stimuli and inhibitors. The data set allowed for an algorithmically driven assembly of related cell types defined by surface antigen expression, providing a superimposable map of cell signaling responses in combination with drug inhibition. Visualized in this manner, the analysis revealed previously unappreciated instances of both precise signaling responses that were bounded within conventionally defined cell subsets and more continuous phosphorylation responses that crossed cell population boundaries in unexpected manners yet tracked closely with cellular phenotype. Collectively, such single-cell analyses provide system-wide views of immune signaling in healthy human hematopoiesis, against which drug action and disease can be compared for mechanistic studies and pharmacologic intervention.

Single-cell phospho-specific flow cytometric analysis demonstrates biochemical and functional heterogeneity in human hematopoietic stem and progenitor compartments.

The low frequency of hematopoietic stem and progenitor cells (HSPCs) in human BM has precluded analysis of the direct biochemical effects elicited by cytokines in these populations, and their functional consequences. Here, single-cell phospho-specific flow cytometry was used to define the signaling networks active in 5 previously defined human HSPC subsets. This analysis revealed that the currently defined HSC compartment is composed of biochemically distinct subsets with the ability to respond rapidly and directly in vitro to a broader array of cytokines than previously appreciated, including G-CSF. The G-CSF response was physiologically relevant-driving cell-cycle entry and increased proliferation in a subset of single cells within the HSC compartment. The heterogeneity in the single-cell signaling and proliferation responses prompted subfractionation of the adult BM HSC compartment by expression of CD114 (G-CSF receptor). Xenotransplantation assays revealed that HSC activity is significantly enriched in the CD114(neg/lo) compartment, and almost completely absent in the CD114(pos) subfraction. The single-cell analyses used here can be adapted for further refinement of HSPC surface immunophenotypes, and for examining the direct regulatory effects of other factors on the homeostasis of stem and progenitor populations in normal or diseased states.