|
|
Blood, Vol. 113, Issue 3, 635-645, January 15, 2009
Gene expression profiling of pulmonary mucosa-associated lymphoid tissue lymphoma identifies new biologic insights with potential diagnostic and therapeutic applications
Blood Chng et al.
113: 635
Supplemental materials for: Chng et al
Supplemental Methods Network Analysis
To assess possible interactions between differentially expressed genes, we performed pathway / network analysis using a Web-based software tool, MetaCore™ (GeneGo Inc, St Joseph, MI, USA). MetaCore™ contains an interactive, manually annotated database derived from literature publications on proteins and small molecules that allows for representation of biological functionality and integration of functional, molecular, or clinical information.1 Several algorithms to enable both the construction and analysis of gene networks are integrated as previously described. The output p-values reflect scoring, prioritization and statistical significance of networks according to relevance of input data. Gene Set Enrichment Analysis (GSEA)
GSEA has been described elsewhere.2 There are three main elements to the GSEA methods. First an enrichment score (ES) that reflects the degree to which a gene set S is overrepresented at the extremes (top or bottom) of the entire ranking list L. A weighted Kolmogorov-Smirnov–like statistic is used. The statistical significance (norminal p-value) of the ES is estimated by using an empirical phenotype-based permutation test procedure that preserves the complex correlation structure of the gene expression data. Specifically, a null distribution of the ES was generated by permutating the phenotype labels and recomputing the ES of the gene set for the permutated labels. The nominal p-value of the observed ES is then calculated relative to this null distribution. When an entire database of gene sets is evaluated, the estimated significance level is adjusted to account for multiple hypothesis testing. The ES is first normalized for each gene set to account for the size of the set. Proportion of false positives is controlled by calculated the FDR which is computed by comparing the tails of the observed and null distribution for the normalized ES. The goal of GSEA is to determine whether the members of S are randomly distributed throughout L or primarily found at the top or bottom, in which case the gene set is correlated with the phenotypic class distinction. The ranking metric used was Signal2Noise, and the phenotype was permutated with 1000 permutations to estimate the statistical significance of enrichment. FDR was controlled at 5%. The C1: position, C2: curated, and C3: promoter motifs gene-sets were analyzed. REFERENCES 1. Nikolsky Y, Ekins S, Nikolskaya T, Bugrim A. A novel method for generation of signature networks as biomarkers from complex high throughput data. Toxicol Lett. 2005;158:20–29.
2. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102:15545–15550.
Files in this Data Supplement:
- Table S1. Tissues included in comparative gene expression analysis (PDF, 11.1 KB)
- Table S2. T-cell genes over-expressed in MALT compared to other B-cell malignancies (PDF, 28.4 KB)
- Table S3. GO analysis of T-cell genes over-expressed in MALT compared to other B-cell malignancies listed in supplemental table 3. (Immune related categories are highlighted in red) (PDF, 11 KB)
- Table S4. 50 genes that is differentially expressed between MALT and other B-cell tumors (PDF, 18.6 KB)
- Table S5. Differentially expressed genes between MALT with and without translocations involving MALT1 (PDF, 58 KB)
- Table S6. Enriched GO biological categories among genes over-expressed in tumors with MALT1 translocations (PDF, 16.8 KB)
- Table S7. Enriched GO biological categories among genes under-expressed in tumors with MALT1 translocations (PDF, 17.2 KB)
- Table S8. Gene-sets enriched in the RGS13 and FKBP11 group of MALT (PDF, 10.9 KB)
- Figure S1. Genes differentially expressed between samples with and without translocation involving MALT1 (JPG, 168 KB)
-
(A) The clustering of genes differentially expressed between MALT lymphomas with and without translocations and the clustering of tumors based on the expression pattern of these genes are represented as a heatmap. Red and green indicate over- and under-expressed genes respectively. The translocation status is indicated by a color bar at the bottom of the heatmap with the legend represented by colored blocks below. The expression of MALT1 is indicated and clearly has the highest expression level in the 3 tumors with t(14;18). The scale of expression used is similar to Fig. 1. When subjected to network analysis, (B) the genes upregulated in MALT1 translocated cases for an extensive interacting network involving the NFKB pathway (enclosed by orange box), JAK-STAT signaling and chemokine induced G-coupled protein signaling. The genes that are differentially expressed between tumors with and without translocations are indicated by red targets on the gene icon. (C) On the other hand, downregulated genes are enriched for plasma cell related genes (left panel) as well as a small network involve in MAPK-JNK signaling (right panel).
- Figure S2. Correlation between strongly correlated gene expression and lung contamination or sex (JPG, 141 KB)
-
Heatmap representing the unsupervised clustering of MALT samples based on genes variably expressed across these tumors. The first gene cluster represents genes related to lung tissue, monocytes and erythrocytes and samples with high expression of these genes are those with the highest percentage of lung tissue in the biopsy specimen therefore reflecting contamination signature from lung tissue. Another strong gene signature is that of genes located on Y chromosome. The tumors over-expressing these genes are actually from male patients. In subsequent analysis (Fig. 4), these 2 gene clusters were removed as they are not biologically relevant and may affect sample clustering.
- Figure S3. Genes with spiked expression segregate into distinct networks that integrate through c-SRC (JPG, 79.1 KB)
-
The genes with spiked expression form 2 separate interacting networks, one involving NFKB signaling and apoptosis pathway (highlighted in green) and the other involving G-protein signaling and oncogenic receptor signaling (highlighted in purple). Interestingly, the spiked genes constituting these 2 networks are also segregated according to their translocations status (Fig. 5).
|
|
