Thunder Data Portal
Introduction
Technical Details
Data Downloads
Interesting genes
Show all genes
|
| |
Index of Interesting Genes
This page provides links to list of alternatively spliced genes identified in the Wu et al., manuscript, as well as genes identified to be differentially expressed between the conditions at the gene level.
Alternatively spliced genes identified in Wu et al.
343 genes exhibited a Thunder genotype x probeset ANOVA p<0.05 in four or more cell types. The highest ranked of these were manually reviewed against mouse mRNAs as illustrated in Figure 5 of the original manuscript, and were placed into one of six categories (groups A-F).
- Group A: Genes with known alt. splicing enhanced in Thunder mutant. Genes with Hnrpll-dependent probesets corresponding to known alternative spliced internal exons in UCSC reference mouse mRNAs were assigned to Group A if the signal increased in thu/thu cells, implying a failure of silencing in the mutant T cells (as occurs for Ptprc exons 4-6).
- Group B: Genes with known alt. splicing inhibited in Thunder mutant. Genes with Hnrpll-dependent probesets corresponding to known alternative spliced internal exons in UCSC reference mouse mRNAs were assigned to Group B if the signal decreased in thu/thu cells, implying increased exon inclusion in the mutant cells.
- Group C: Genes with novel alt. splicing enhanced in Thunder mutant. Genes with Hnrpll-dependent probesets not currently identified as having alternatively spliced exons in reference mouse mRNAs were assigned to Group C if the signal increased in thu/thu cells, implying a failure of silencing in the mutant T cells.
- Group D: Genes with novel alt. splicing inhibited in Thunder mutant. Genes with Hnrpll-dependent probesets not currently identified as having alternatively spliced exons in reference mouse mRNAs were assigned to Group D if the signal decreased in thu/thu cells, implying increased exon inclusion in the mutant cells.
- Group E: Genes with 5' exons altered in Thunder mutant. Genes were assigned to Group E if the Hnrpll-dependent probesets corresponded to first exons of transcripts and therefore potentially represent alternative promoters and start sites. A number of these are supported by transcript, EST and published evidence.
- Group F: Genes with complex splicing patterns altered in Thunder mutant. Genes were assigned to Group F if they exhibited complex patterns of Hnrpll-dependence. Many of these appear to involve changes in overall mRNA expression from single or multi-transcript gene complexes.
- Group Memory: Genes differentially spliced between memory and naive T cells. The data was also queried for alternative splicing occurring in wild-type T cells between naive and memory subsets (Figure 5 - manuscript). Over 500 genes were identified that scored p<0.001 (2-way ANOVA) in both memory cell types (CD4 and CD8).
Genes differentially expressed between conditions
To identify genes with overall expression levels that were significantly affected by genotype (WT or Thunder mutant), memory status (naive or memory) or T cell subset (CD4 or CD8), we used linear modeling and ANOVA. For the case where there were not interactions between factors (for example, the effect of the Thunder mutation did not depend on whether they were CD4 or CD8 T cells , naive or memory), we fit Model 1:
Model 1: Y_ijkl = u + G_i + M_j+C_k + e_ijkl
Where:
- Y_ijkl = log2(expression level) for a particular gene in genotype "i" (i = WT or thu/thu), memory status "j" (j = naive[n] or memory[m]), T cell subset "k" (k = CD4 or CD8), and biological replicate "l" ("l" = 1-4).
- G_i = the Genotype effect
- M_j = the Memory effect
- C_k = the T cell subset effect
- e_ijkl = normally distributed random error term.
To identify genes with interactions between factors (for example, to find genes where the Thunder mutation only affected expression levels in CD8 T cells), we fit modified versions of Model 1 that included the relevant interaction terms.
|
|
