CoChat

SLICE: determining cell differentiation and lineage based on single cell entropy

Open access
Guo, Minzhe, Bao, Erik L., Wagner, Michael, Whitsett, Jeffrey A., Xu, Yan · 2016 · Nucleic Acids Research

The DOI matches its Crossref record; 34 of the 34 references it cites were found in a registry or an index.

Abstract

A complex organ contains a variety of cell types, each with its own distinct lineage and function. Understanding the lineage and differentiation state of each cell is fundamentally important for the ultimate delineation of organ formation and function. We developed SLICE, a novel algorithm that utilizes single-cell RNA-seq (scRNA-seq) to quantitatively measure cellular differentiation states based on single cell entropy and predict cell differentiation lineages via the construction of entropy directed cell trajectories. We validated our approach using three independent data sets with known lineage and developmental time information from both Homo sapiens and Mus musculus. SLICE successfully measured the differentiation states of single cells and reconstructed cell differentiation trajectories that have been previously experimentally validated. We then applied SLICE to scRNA-seq of embryonic mouse lung at E16.5 to identify lung mesenchymal cell lineage relationships that currently remain poorly defined. A two-branched differentiation pathway of five fibroblastic subtypes was predicted using SLICE. The present study demonstrated the general applicability and high predictive accuracy of SLICE in determining cellular differentiation states and reconstructing cell differentiation lineages in scRNA-seq analysis.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/nar/gkw1278

Cites

34 of 34 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1. The design of the SLICE algorithm. (A) The schematic flow of the SLICE algorithm. (B) The schematic flow of scEntropy.
Figure 1. The design of the SLICE algorithm. (A) The schematic flow of the SLICE algorithm. (B) The schematic flow of scEntropy.CC BY-NC · PubMed Central
Figure 2. Single cell entropy (scEntropy) measures differentiation states of individual cells. (A) The scEntropies of lung alveolar type 2 (AT2, n = 101) cells significantly decreased along with mouse lung development. The cells were isolated from E14.5, E16.5, E18.5 and adult mouse lung (13). (B) The decrease of scEntropies of AT2 cells correlates with the progression of AT2 differentiation. The expression patterns of early progenitor markers (Sox9 and Sox11) and mature AT2 markers (Sftpb and Sftpc) were used to validate the order of scEntropies. (C) The scEntropies of human skeletal muscle myoblast (HSMM, n = 223) cells significantly decreased during myoblast differentiation. The assignments of cells to proliferating cells (n = 96) and differentiating myoblast cells (n = 127) were pre-defined by Trapnell et al. (7). (D) The scEntropies of human early embryonic (HEE, n = 88) cells significantly decreased during human early embryo development. The cell developmental stages were defined by Yan et al. (14). E) The scEntropy measurement of the predicted fibroblast subtypes from E16.5 mouse lung (FB, n = 79). Fibroblast subtypes were defined by Guo et al. (16); PMP: proliferative mesenchymal progenitor, IF1: intermediate fibroblast 1, IF2: intermediate fibroblast 2, MyoF: myofibroblast/smooth muscle-like, MFB: matrix fibroblast. In (B), (C), (D) and (E), orange lines represent the polynomial fits of degree 3 of scEntropies using least squares regression and P-values were obtained using the Analysis of Variance (ANOVA), with P-value <0.0001 as the threshold for significance.
Figure 2. Single cell entropy (scEntropy) measures differentiation states of individual cells. (A) The scEntropies of lung alveolar type 2 (AT2, n = 101) cells significantly decreased along with mouse lung development. The cells were isolated from E14.5, E16.5, E18.5 and adult mouse lung (13). (B) The decrease of scEntropies of AT2 cells correlates with the progression of AT2 differentiation. The expression patterns of early progenitor markers (Sox9 and Sox11) and mature AT2 markers (Sftpb and Sftpc) were used to validate the order of scEntropies. (C) The scEntropies of human skeletal muscle myoblast (HSMM, n = 223) cells significantly decreased during myoblast differentiation. The assignments of cells to proliferating cells (n = 96) and differentiating myoblast cells (n = 127) were pre-defined by Trapnell et al. (7). (D) The scEntropies of human early embryonic (HEE, n = 88) cells significantly decreased during human early embryo development. The cell developmental stages were defined by Yan et al. (14). E) The scEntropy measurement of the predicted fibroblast subtypes from E16.5 mouse lung (FB, n = 79). Fibroblast subtypes were defined by Guo et al. (16); PMP: proliferative mesenchymal progenitor, IF1: intermediate fibroblast 1, IF2: intermediate fibroblast 2, MyoF: myofibroblast/smooth muscle-like, MFB: matrix fibroblast. In (B), (C), (D) and (E), orange lines represent the polynomial fits of degree 3 of scEntropies using least squares regression and P-values were obtained using the Analysis of Variance (ANOVA), with P-value <0.0001 as the threshold for significance.CC BY-NC · PubMed Central
Figure 3. SLICE reconstructed the differentiation lineage of mouse lung alveolar type 2 (AT2) cells. (A) Principal component analysis (PCA) of AT2 cells (n = 101) isolated from E14.5, E16.5, E18.5 and adult mouse lung. Genes (n = 2973) expressed in at least 30% of the cells and with a non-zero variance in log2-transformed expression were used in the PCA analysis. (B) Four cell clusters and their stable states were identified by SLICE. (C) The lineage model of AT2 cells during lung development inferred by SLICE. (D) The reconstruction of transitional path from C1→C2 →C3→C4 of AT2 cells. The path was reconstructed by concatenating pairwise shortest-paths between successive stable states (C1, C2, C3 and C4) in the minimum spanning tree of the cells. (E) The expression patterns of known markers indicated that the reconstructed trajectory represented the progression of AT2 cell differentiation. In (A)–(D), sizes of cells were proportional to the max-normalized scEntropies of the cells, and dark-gray nodes represent the detected stable states.
Figure 3. SLICE reconstructed the differentiation lineage of mouse lung alveolar type 2 (AT2) cells. (A) Principal component analysis (PCA) of AT2 cells (n = 101) isolated from E14.5, E16.5, E18.5 and adult mouse lung. Genes (n = 2973) expressed in at least 30% of the cells and with a non-zero variance in log2-transformed expression were used in the PCA analysis. (B) Four cell clusters and their stable states were identified by SLICE. (C) The lineage model of AT2 cells during lung development inferred by SLICE. (D) The reconstruction of transitional path from C1→C2 →C3→C4 of AT2 cells. The path was reconstructed by concatenating pairwise shortest-paths between successive stable states (C1, C2, C3 and C4) in the minimum spanning tree of the cells. (E) The expression patterns of known markers indicated that the reconstructed trajectory represented the progression of AT2 cell differentiation. In (A)–(D), sizes of cells were proportional to the max-normalized scEntropies of the cells, and dark-gray nodes represent the detected stable states.CC BY-NC · PubMed Central
Figure 4. SLICE reconstructed the differentiation lineage of single cells from human skeletal muscle myoblasts. (A) Human skeletal muscle myoblast (HSMM) cells (n = 266) in the two-dimensional space derived from a principal component analysis using a set of marker genes (Supplementary Data). Cell states (indicated by three distinct colours) were defined by the original analysis (7). (B) Four cell clusters and their stable states identified by SLICE. (C) The lineage model of HSMM cells inferred by SLICE. (D) The reconstruction of transitional path from cluster C1, through cluster C2, to cluster C3 of HSMM cells. The path was reconstructed by connecting pairwise shortest-paths between successive stable states (C1, C2 and C3) in the cell-cell network. (E) The expression patterns of myogenic differentiation markers along the reconstructed transitional path (left) were consistent with the patterns based on pseudotime (right) using Monocle (7). In (A)–(D), sizes of cells were proportional to the max-normalized scEntropies of the cells, and dark-grey nodes represent the detected stable states.
Figure 4. SLICE reconstructed the differentiation lineage of single cells from human skeletal muscle myoblasts. (A) Human skeletal muscle myoblast (HSMM) cells (n = 266) in the two-dimensional space derived from a principal component analysis using a set of marker genes (Supplementary Data). Cell states (indicated by three distinct colours) were defined by the original analysis (7). (B) Four cell clusters and their stable states identified by SLICE. (C) The lineage model of HSMM cells inferred by SLICE. (D) The reconstruction of transitional path from cluster C1, through cluster C2, to cluster C3 of HSMM cells. The path was reconstructed by connecting pairwise shortest-paths between successive stable states (C1, C2 and C3) in the cell-cell network. (E) The expression patterns of myogenic differentiation markers along the reconstructed transitional path (left) were consistent with the patterns based on pseudotime (right) using Monocle (7). In (A)–(D), sizes of cells were proportional to the max-normalized scEntropies of the cells, and dark-grey nodes represent the detected stable states.CC BY-NC · PubMed Central
Figure 5. SLICE reconstructed the differentiation lineage of single cells from human early embryo (HEE) development. (A) HEE cells (n = 88) with development stage information in the two-dimensional space derived from a principal component analysis using genes (n = 10 601) expressed in at least 30% of the cells and with a non-zero variance in log2-transformed expression. (B) Four cell clusters and their stable states identified by SLICE. (C) The inferred lineage model of HEE cells during human early embryo development. (D) The reconstruction of transitional path from cluster C1, through C2 and C3, to C4 of HEE cells. The path was reconstructed by concatenating pairwise shortest-paths between successive stable states (C1, C2, C3 and C4) in the minimum spanning tree of cells. (E) The expression patterns of markers along the inferred transitional path (left) were consistent with the stages defined in Yan et al. (14). TE, trophectoderm; PE, primitive endoderm; EPI, epiblast. In (A)–(D), sizes of cells were proportional to the max-normalized scEntropies of the cells, and dark-grey nodes represent the detected stable states.
Figure 5. SLICE reconstructed the differentiation lineage of single cells from human early embryo (HEE) development. (A) HEE cells (n = 88) with development stage information in the two-dimensional space derived from a principal component analysis using genes (n = 10 601) expressed in at least 30% of the cells and with a non-zero variance in log2-transformed expression. (B) Four cell clusters and their stable states identified by SLICE. (C) The inferred lineage model of HEE cells during human early embryo development. (D) The reconstruction of transitional path from cluster C1, through C2 and C3, to C4 of HEE cells. The path was reconstructed by concatenating pairwise shortest-paths between successive stable states (C1, C2, C3 and C4) in the minimum spanning tree of cells. (E) The expression patterns of markers along the inferred transitional path (left) were consistent with the stages defined in Yan et al. (14). TE, trophectoderm; PE, primitive endoderm; EPI, epiblast. In (A)–(D), sizes of cells were proportional to the max-normalized scEntropies of the cells, and dark-grey nodes represent the detected stable states.CC BY-NC · PubMed Central
Figure 6. SLICE predicted a two-branched fibroblast cell differentiation pathway at E16.5 mouse lung. (A) Principal component analysis of five fibroblast (FB) subtype cells (n = 79) using a set of highly variable genes (n = 252, Supplementary Data). Our previous analyses of single cell RNA-seq data from E16.5 mouse lung identified five distinct fibroblast subtypes, termed as PMP, MyoF, MFB and two unknown fibroblast (IF1 and IF2). Their hierarchical relationships are unknown. (B) Four cell clusters and their stable states identified by SLICE. (C) A two-branched lineage model of FB cells at E16.5 mouse lung inferred by SLICE. (D) and (E) showed the reconstructed cell transitional path for each branch, respectively. Each path was reconstructed by concatenating pairwise shortest-paths between successive stable states in the minimum spanning tree of cells. (F) A schematic presentation of the inferred two-branched differentiation pathway among the five predicted FB subtypes. (G) The expression patterns of markers along the two transitional paths ‘C1→C3→C4’ (left) and ‘C1→C2’ (right) suggested a two-branched cell differentiation pathway of lung fibroblasts at E16.5 (i.e. PMP→MFB and PMP→MyoF). (H) Temporal patterns of differentially expressed genes in transitional path ‘C1→C3→C4’ (top) and in transitional path ‘C1→C2’ (bottom). In (H), temporal pattern 1 mainly consisted of differentially down-expressed genes while temporal pattern 3 is primarily comprised of differentially up-expressed genes. (I) Enriched functions of genes (n = 493) induced in transitional path ‘C1→C3→C4’ (top) including ‘extracelluar matrix’ and ‘collagen and elastic fiber formation’ which were closely associated with the functions of matrix fibroblasts. Enriched functions of genes (n = 622) induced in transitional path ‘C1→C2’ (bottom) including ‘muscle development and differentiation’ and ‘smooth muscle contraction’. In (A)–(E), sizes of cells were proportional to the max-normalized scEntropies of the cells, and dark-grey nodes represent the detected stable states.
Figure 6. SLICE predicted a two-branched fibroblast cell differentiation pathway at E16.5 mouse lung. (A) Principal component analysis of five fibroblast (FB) subtype cells (n = 79) using a set of highly variable genes (n = 252, Supplementary Data). Our previous analyses of single cell RNA-seq data from E16.5 mouse lung identified five distinct fibroblast subtypes, termed as PMP, MyoF, MFB and two unknown fibroblast (IF1 and IF2). Their hierarchical relationships are unknown. (B) Four cell clusters and their stable states identified by SLICE. (C) A two-branched lineage model of FB cells at E16.5 mouse lung inferred by SLICE. (D) and (E) showed the reconstructed cell transitional path for each branch, respectively. Each path was reconstructed by concatenating pairwise shortest-paths between successive stable states in the minimum spanning tree of cells. (F) A schematic presentation of the inferred two-branched differentiation pathway among the five predicted FB subtypes. (G) The expression patterns of markers along the two transitional paths ‘C1→C3→C4’ (left) and ‘C1→C2’ (right) suggested a two-branched cell differentiation pathway of lung fibroblasts at E16.5 (i.e. PMP→MFB and PMP→MyoF). (H) Temporal patterns of differentially expressed genes in transitional path ‘C1→C3→C4’ (top) and in transitional path ‘C1→C2’ (bottom). In (H), temporal pattern 1 mainly consisted of differentially down-expressed genes while temporal pattern 3 is primarily comprised of differentially up-expressed genes. (I) Enriched functions of genes (n = 493) induced in transitional path ‘C1→C3→C4’ (top) including ‘extracelluar matrix’ and ‘collagen and elastic fiber formation’ which were closely associated with the functions of matrix fibroblasts. Enriched functions of genes (n = 622) induced in transitional path ‘C1→C2’ (bottom) including ‘muscle development and differentiation’ and ‘smooth muscle contraction’. In (A)–(E), sizes of cells were proportional to the max-normalized scEntropies of the cells, and dark-grey nodes represent the detected stable states.CC BY-NC · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 34 references

Research works printed but not linked: 2

  1. 10.2307/2289936 · DOI
  2. Shin J. Berg D.A. Zhu Y. Shin J.Y. Song J. Bonaguidi M.A. Enikolopov G. Nauen D.W. Christian K.M. Ming G.L. . Single-cell RNA-Seq with waterfall reveals molecular cascades underlying adult neurogenesis. Cell Stem Cell 2015. · DOI · Paper page

Cited by

See all 185

Metadata

Authors
Guo, M., Bao, E. L., Wagner, M., Whitsett, J. A., & Xu, Y.
Affiliations
  • Cincinnati Children's Hospital Medical Center (US)
  • Harvard University (US)
Corresponding author
  • Guo, Minzhe
  • Bao, Erik L.
  • Xu, Yan
Venue
Nucleic Acids Research
Published
2016-12-19
Volume
45
Issue
7
Pages
gkw1278
ISSN
  • 0305-1048 (ISSN-L)
  • 1362-4962
OpenAlex
W2561394844
Cited by
178
References
34
Citations by year
2026: 1 · 2025: 20 · 2024: 28 · 2023: 21 · 2022: 29 · 2021: 14 · 2020: 24 · 2019: 15 · 2018: 19 · 2017: 7
FWCI
7.10
Type
journal article
Language
en
Open access
Locations
  • Nucleic Acids Research (publishedVersion, cc-by-nc)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • Digital Access to Scholarship at Harvard (DASH) (Harvard University) (submittedVersion, cc-by)
Funders
  • National Institutes of Health
  • National Heart, Lung, and Blood Institute (R01HL095580, U01HL122642, R01HL136722)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene Regulatory Network Analysis
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Developmental Biology and Gene Regulation
Keywords
cell differentiation, cell lineage, single-cell RNA sequencing, fibroblast subtypes
MeSH terms
Algorithms, Animals, Cell Differentiation, Embryo, Mammalian, Fibroblasts, Humans, Lung, Sequence Analysis, RNA, Cell Lineage, Gene Expression Profiling, Myoblasts, Skeletal, Embryonic Development, Mice, Alveolar Epithelial Cells, Single-Cell Analysis

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: SLICE: determining cell differentiation and lineage based on single cell entropy. https://app.cochat.ai/papers/10.1093/nar/gkw1278
The paper
Guo, M., Bao, E. L., Wagner, M., Whitsett, J. A., & Xu, Y. (2016). SLICE: determining cell differentiation and lineage based on single cell entropy. Nucleic Acids Research. https://doi.org/10.1093/nar/gkw1278

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.