Skip to main page content
U.S. flag
See More

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2013 Mar 7:14:89.
doi: 10.1186/1471-2105-14-89.

Digital sorting of complex tissues for cell type-specific gene expression profiles

Yi Zhong  1 , Ying-Wooi Wan, Kaifang Pang, Lionel M L Chow, Zhandong Liu
Affiliations

Digital sorting of complex tissues for cell type-specific gene expression profiles

Yi Zhong et al. BMC Bioinformatics. .

Abstract

Background: Cellular heterogeneity is present in almost all gene expression profiles. However, transcriptome analysis of tissue specimens often ignores the cellular heterogeneity present in these samples. Standard deconvolution algorithms require prior knowledge of the cell type frequencies within a tissue or their in vitro expression profiles. Furthermore, these algorithms tend to report biased estimations.

Results: Here, we describe a Digital Sorting Algorithm (DSA) for extracting cell-type specific gene expression profiles from mixed tissue samples that is unbiased and does not require prior knowledge of cell type frequencies.

Conclusions: The results suggest that DSA is a specific and sensitivity algorithm in gene expression profile deconvolution and will be useful in studying individual cell types of complex tissues.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Unbiased estimation of tissue type specific profiles. (a) Mixing proportions were estimated using markers for liver, brain and lung. DSA estimation can recapitulate the true percentage of each cell type in the mixture. (b-d) DSA estimation of liver, brain and lung gene expression profiles compared against true expression profiles measured using pure tissue samples. (e-f) ROC analysis on differential gene expression analysis of brain vs. liver and lung vs. liver using DSA.
Figure 2
Figure 2
Unbiased estimation of cell type specific profiles. (a) Cell type frequencies were estimated using markers for IM-9 cells (green), Raji cells (blue), Jurkat cells (red), and THP-1 cells (purple). (b) DSA estimation of the gene expression profiles of IM-9 cells compared against true expression profiles measured using pure cell samples. (c) ROC analysis on differential gene expression analysis of estimated IM-9 cells vs. Jurkat cells using DSA.
Figure 3
Figure 3
Comparison between DSA and PSEA. (a) Fold change estimated by DSA compared against the true fold change between liver and brain samples. The dotted line represents the reference line where all the points should follow. (b) Fold change estimated by PSEA compared against the true fold change between liver and brain samples.
Figure 4
Figure 4
The estimated transcriptomes for 6 different immune cell types were plotted against the gene expression measured on arrays. Cell types that have higher percentage in the tissue sample tend to have better estimation accuracy. (a-f) Scatter plots of estimated profile against microarray measures in Eosinophil, Myeloid Dendritic, Mature B-cells, Granulocyte, Naïve B-cells, and Basophils.
Figure 5
Figure 5
(a-c) The AUC analysis of cell types that have high and low confidence of deconvolved gene expression profile. Eosinophil and myeloid dendritic cells have the best AUC scores since these two cell types have the highest proportions in the mixed samples. Naïve B-cells and basophils yield poor but still informative AUC scores, as these two cell types have the lowest frequency in the mixed samples. (d) The plot of mean square of error (MSE) and weight against signal-to-noise ratio (SNR). The best cut-off point was observed around 45. Cell types that are present at too low of a frequency in a given tissue will have dramatically increased errors.
Figure 6
Figure 6
(a) The percentage of TAMs in Hodgkin’s lymphoma tumors was negatively associated with progression-free survival. (b) Genes that are highly expressed in DSA extracted TAMs are enriched for biological processes characteristic of macrophages, such as response to wounding, immune, inflammatory and defense response.
See this image and copyright information in PMC

Comment in

References

    1. Cobb JP. Application of genome-wide expression analysis to human health and disease. Proc Natl Acad Sci. 2005;102:4801–4806. doi: 10.1073/pnas.0409768102. - DOI - PMC - PubMed
    1. Repsilber D, Kern S, Telaar A, Walzl G, Black GF, Selbig J, Parida SK, Kaufmann SH, Jacobsen M. Biomarker discovery in heterogeneous tissue samples -taking the in-silico deconfounding approach. BMC Bioinformatics. 2010;11:27. doi: 10.1186/1471-2105-11-27. - DOI - PMC - PubMed
    1. Palmer C, Diehn M, Alizadeh AA, Brown PO. BMC Genomics. 2006. p. 115. - DOI - PMC - PubMed
    1. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144:646–674. doi: 10.1016/j.cell.2011.02.013. - DOI - PubMed
    1. Yang F. Stromal expression of connective tissue growth factor promotes angiogenesis and prostate cancer tumorigenesis. Cancer Res. 2005;65:8887–8895. doi: 10.1158/0008-5472.CAN-05-1702. - DOI - PubMed

Publication types

MeSH terms