Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis

By Jacob H. Levine, Erin F. Simonds, Sean Bendall, K. Davis, E. Amir, Michelle D. Tadmor, O. Litvin, Harris G. Fienberg, Astraea Jager, E. Zunder and 6 more. 2015

2,148 citations recorded by the source corpus.

SUMMARY Acute myeloid leukemia (AML) manifests as phenotypically and functionally diverse cells, often within the same patient. Intratumor phenotypic and functional heterogeneity have been linked primarily by physical sorting experiments, which assume that functionally distinct subpopulations can be prospectively isolated by surface phenotypes. This assumption has proven problematic and we therefore developed a data-driven approach. Using mass cytometry, we profiled surface and intracellular signaling proteins simultaneously in millions of healthy and leukemic cells. We developed PhenoGraph, which algorithmically defines phenotypes in high-dimensional single-cell data. PhenoGraph revealed that the surface phenotypes of leukemic blasts do not necessarily reflect their intracellular state. Using hematopoietic progenitors, we defined a signaling-based measure of cellular phenotype, which led to isolation of a gene expression signature that was predictive of survival in independent cohorts. This study presents new methods for large-scale analysis of single-cell heterogeneity and demonstrates their utility, yielding insights into AML pathophysiology.

Subfields: Cancer Research, Hematology, Oncology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cancer Research

In Hematology

In Oncology

Citations in the corpus

Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets (2019, methodology): “Clustering algorithms that directly interrogate high-dimensional data, such as FlowSOM 11 and PhenoGraph 12 , are often used in conjunction with 2D maps to present annotated cell clusters to the viewer.”
Prioritizing network communities (2018, methodology): “For example, gene communities detected in a gene interaction network 11 provide predictions/hypotheses about disease pathways 2,3 , but to con ﬁ rm these predictions scientists have to test every detected community by performing experiments in a wet laboratory 3,8 .”
SCANPY: large-scale single-cell gene expression data analysis (2018, methodology): “com/jacoblevine/ PhenoGraph) [19], the single-cell experiment design tool MIMOSCA (https://github.”
Distinct cellular mechanisms underlie anti-CTLA-4 and anti-PD-1 checkpoint blockade (2017, positive): “Using a metaclustering approach, in which populations first identified at the individual mouse level using PhenoGraph are then allowed to merge across cohorts (STAR Methods), 14 T cell populations were identified.”
SOM-based embedding improves efficiency of high-dimensional cytometry data analysis (2019, methodology): “e datasets were pre-processed as in the original article [13], ungated cells were excluded from benchmarking.”
COMPREHENSIVE CLASSIFICATION OF RETINAL BIPOLAR NEURONS BY SINGLE-CELL TRANSCRIPTOMICS (2016, positive): “Two graph clustering algorithms, Louvain-Jaccard (Blondel et al., 2008; Levine et al., 2015) and Infomap (Rosvall and Bergstrom, 2008), exhibited superior performance as judged by a post hoc comparison of predicted clusters to known BC types, their ability to resolve clusters when applied to…”
Single-Cell Profiling and SCOPE-Seq Reveal Lineage Dynamics of Adult Ventricular-Subventricular Zone Neurogenesis and NOTUM as a Key Regulator (2020, methodology): “REAGENT or RESOURCE SOURCE IDENTIFIER Mouse: ratNes::FLPOER The Joyner lab Wojcinski et al., 2017 Mouse: Rosa26FRT-STOP-FRT-TdTomato The Jackson Laboratory JAX: 021875 Mouse: Nestin-CFP The Joyner lab Wojcinski et al., 2017 Oligonucleotides Template Switching Oligonucleotide AAGCAGTGGTATCAACGCAGAGTG”
Assessment of computational methods for the analysis of single-cell ATAC-seq data (2019, methodology): “For this study we used three commonly used clustering methods: k-means, hierarchical 1223 clustering (with default ward linkage) as implemented in the scikit-learn library [31] 1224 and Louvain clustering (a community-detection-based method) [32, 33] as implemented 1225 in Scanpy [34], For both hier”
Characterization of cell fate probabilities in single-cell data with Palantir (2019, positive): “Phenograph clustering of data revealed a cluster of cells with low molecule count distribution, which was excluded from the analysis.”
Current best practices in single‐cell RNA‐seq analysis: a tutorial (2019, methodology): “A commonly used normalization method for full-length scRNA-seq data is TPM normalization (Li et al , 2009), which comes from bulk RNA-seq analysis.”
Interpretable dimensionality reduction of single cell transcriptome data with deep generative models (2017, methodology): “Technical developments have made it possible to measure the DNA and/or RNA molecules in single cells by single-cell sequencing 4 – 15 or protein content by ﬂ ow or mass cytometry 16, 17 .”
OMIP-044: 28-color immunophenotyping of the human dendritic cell compartment (2018, methodology): “Given that exhaustive analysis using manual two-dimensional gating is not feasible (24,25), several novel approaches for exploratory analysis of high-dimensional single-cell data sets have been reported in the past years, among them t-stochastic neighbor embedding (t-SNE) (26,27), which can be used ”