In the transcriptome comparison, rat brain tissue is more similar to mouse brain than rat heart (Figure 1). Now, I have the gene expression of each tissue in two species (Figure 2), how can I cluster these two data? Ultimately I want to look like Figure 1.
- Maybe I need to find out the orthologous gene first?
- I think I can use "dist" and "hclust" in r, but now one of the gene lists is for rat and the other is for mouse.
- And I have already normalized the gene expression.
Thank you so much.
•
written
by
cherhorowitz •
2 answers
Hi, Yes, you need to find the gene mapping between species if you want to plot them together. Be aware of the batch effect; tissues should be clustered together. see https://f1000research.com/articles/4-121/v1
All are covered in my book https://divingintogeneticsandgenomics.ck.page/products/cell-line-to-command-line
You can find the gene-to-gene mappings in ensembl biomart. Best to use only one-to-one single-copy orthologues.
Log in to answer this question.
-
Weird edgeR normalization results for one tissue
written by imetrev •
-
Tutorial: Perform Hierarchical Clustering in R with ggdendro Visualization
written by Jeremy •
-
Expression differs when running plotTranscripts vs boxplot of FPKM or coverage values (R)?
written by boone5jones •
-
Differential expression to compare genes between brain regions (Allen Mouse Brain Atlas)
written by dekel •
-
Differences between gene expression expressed as z-score or log2 values in Allen Brain Atlas
written by c.nunez •
-
R ggplot2 creating separate plots for different subsets of rows
written by moritz.lasse •
-
Gene expression and chromatin state
written by whb •
-
gene expres. simulation with R
written by bioinfonewbie •
-
How to distinguish level of expression in pooled-like samples?
written by ahmad mousavi
-
Should I merge .gtf files from multiple tissues together for cuffmerge?
written by pinakin76 •
