Email updates

Keep up to date with the latest news and content from BMC Genomics and BioMed Central.

Open Access Research article

Cell population-specific expression analysis of human cerebellum

Alexandre Kuhn12*, Azad Kumar1, Alexandra Beilina1, Allissa Dillman1, Mark R Cookson1 and Andrew B Singleton1

Author Affiliations

1 Laboratory of Neurogenetics, National Institute on Aging, National Institutes of Health, Bethesda, MD, USA

2 Current address: Microfluidics Systems Biology, Institute for Materials Research and Engineering, A*STAR, 3 Research Link, Singapore, Singapore, 117602

For all author emails, please log on.

BMC Genomics 2012, 13:610  doi:10.1186/1471-2164-13-610

Published: 12 November 2012

Additional files

Additional file 1:

Table S1. Sample characteristics for gene expression series 1. Series 1 was comprised of 43 samples. The minimal, 1st quartile, median, 3rd quartile and maximal age were 15, 20.5, 38, 48, and 72 years, respectively. Table S2: Sample characteristics for gene expression series 2. Series 2 was comprised of 57 samples. The minimal, 1st quartile, median, 3rd quartile and maximal age were 16, 24, 33, 45 and 58 years, respectively. Table S3: Marker genes used to construct reference signals for PSEA. Two different sets of astrocytic markers were used to generate 2 independent astrocytic reference signals. Set 2 was used to assess the robustness of astrocyte-specific changes detected with set 1. For some genes, we also used the following marker genes to test if expression could be detected in additional minor cell populations (see Results): DES (smooth muscle cells), CSPG4 (pericytes), P4HA1 (fibroblasts), PECAM1 (endothelial cells), CD37 (microglia) [35,46-55]. Table S4: Distribution of gene expression models obtained with PSEA upon statistical model building. G, P, A and O stand for the granular, Purkinje cell, astrocyte and oligodendrocyte reference signals, respectively. The third column indicates the average goodness-of-fit (as mean adjusted R2) for probes assigned a particular statistical model.

Format: PDF Size: 72KB Download file

This file can be viewed with: Adobe Acrobat Reader

Open Data

Additional file 2:

Figure S1. Expression levels of marker genes (left) and corresponding population-specific reference signals (right) for the granule (A), Purkinje (B), astrocytic (C) and oligodendrocytic (D) cell populations. For each row, the left panel shows the (log2) expression of marker genes across all samples. The right panel shows the reference signal obtained by averaging expression of the corresponding marker genes. The standard deviation of reference signals was 0.38 (granule cell), 0.36 (Purkinje cell), 0.55 (astrocyte), 0.8 (oligodendrocyte).

Format: DOC Size: 378KB Download file

This file can be viewed with: Microsoft Word Viewer

Open Data

Additional file 3:

Figure S2. Characterization of gene expression models obtained upon statistical model building, for all genes on the microarray. Genes with better goodness-of-fit (higher adjusted R2) had smaller (relative) intercepts, in line with the hypothesized model for total expression (see Methods). Gray lines show the threshold criteria used for selecting expression models for further consideration (intercept/mean<0.5, adjusted R2>0.5).

Format: DOC Size: 71KB Download file

This file can be viewed with: Microsoft Word Viewer

Open Data

Additional file 4:

Figure S3. Population-specific expression levels and associated p-values (for 5,952 genes passing fit quality criteria). Each volcano plot shows the (normalized) values for a particular model coefficient and corresponding (−log10) p-values (y-axis). Model coefficients were normalized by the average gene expression. A: intercept, B: granule cell-specific expression, C: Purkinje cell-specific expression, D: astrocyte-specific expression, E: oligodendrocyte-specific expression.

Format: DOC Size: 572KB Download file

This file can be viewed with: Microsoft Word Viewer

Open Data

Additional file 5:

Table S5. Genes specifically associated with the granule cell reference signal. We selected genes with p-values for granule cell-specific expression at least 1000 times smaller than for any of the other 3 populations, and p-values for expression in these three populations greater than 0.001. Expr and p stand for coefficients (i.e. population-specific expression levels) and associated p-values, respectively. A, O, G and P indicate the corresponding reference signal: astrocyte, oligodendrocyte, granular and Purkinje, respectively.

Format: TXT Size: 401KB Download file

Open Data

Additional file 6:

Table S6. Genes specifically associated with the Purkinje cell reference signal. We selected genes with p-values for Purkinje cell-specific expression at least 1000 times smaller than for any of the other 3 populations, and p-values for expression in these three populations greater than 0.001. Expr and p stand for coefficients (i.e. population-specific expression levels) and associated p-values, respectively. A, O, G and P indicate the corresponding reference signal: astrocyte, oligodendrocyte, granular and Purkinje, respectively.

Format: TXT Size: 16KB Download file

Open Data

Additional file 7:

Table S7. Genes specifically associated with the astrocyte reference signal. We selected genes with p-values for astrocyte-specific expression at least 1000 times smaller than for any of the other 3 populations, and p-values for expression in these three populations greater than 0.001. Expr and p stand for coefficients (i.e. population-specific expression levels) and associated p-values, respectively. A, O, G and P indicate the corresponding reference signal: astrocyte, oligodendrocyte, granular and Purkinje, respectively.

Format: TXT Size: 16KB Download file

Open Data

Additional file 8:

Table S8. Genes specifically associated with the oligodendrocyte reference signal. We selected genes with p-values for oligodendrocyte-specific expression at least 1000 times smaller than for any of the other 3 populations, and p-values for expression in these three populations greater than 0.001. Expr and p stand for coefficients (i.e. population-specific expression levels) and associated p-values, respectively. A, O, G and P indicate the corresponding reference signal: astrocyte, oligodendrocyte, granular and Purkinje, respectively.

Format: TXT Size: 17KB Download file

Open Data

Additional file 9:

Figure S4. Predicted (log2) Purkinje cell-specific expression levels (x-axis) versus (log2) expression levels measured in experimentally isolated cells (y-axis) for all genes that obtained a non-negative, significant (p<0.05) Purkinje cell expression component by PSEA.

Format: DOC Size: 99KB Download file

This file can be viewed with: Microsoft Word Viewer

Open Data

Additional file 10:

Compressed (gzip) archive file (tar) containing computer code used for the PSEA analysis presented here. The archive contains the main R code (psea.R) and associated functions (psea_f.R), sample information (param.txt), subject phenotypes (phenotypeInfo.txt) and microarray probe annotation file (GPL6104-20626.txt).

Format: TGZ Size: 3.5MB Download file

Open Data

Additional file 11:

Compressed (zip) text file with the raw gene expression data used for the main analysis (i.e. subset of samples deposited in GEO GSE15745).

Format: ZIP Size: 11.3MB Download file

Open Data

Additional file 12:

Compressed (zip) text file with the raw gene expression data used for validation (i.e. subset of samples deposited in GEO GSE15745).

Format: ZIP Size: 15.9MB Download file

Open Data