Question

Gene classification in TissueEnrich

0

Entering edit mode

xiaoxiaoxiaobu • 0

@xiaoxiaoxiaobu-24129

Last seen 4.2 years ago

Hello,

I am a little confused with the classification of genes in TissueEnrich:

Tissue Enriched: Genes with an expression level greater than or equal to 1 (TPM or FPKM) that also have at least five-fold higher expression levels in a particular tissue compared to all other tissues.

Another:

Tissue Enhanced: Genes with an expression level greater than or equal to 1 (TPM or FPKM) that also have at least five-fold higher expression levels in a particular tissue compared to the average levels in all other tissues, and that are not considered Tissue Enriched or Group Enriched.

These two classifications looks quite similar. Can anyone please list some examples of these two kind of gens?

Many thanks!

TissueEnrich • 1.3k views

ADD COMMENT • link updated 4.2 years ago by Ashish Jain ▴ 10 • written 4.2 years ago by xiaoxiaoxiaobu • 0

score 0 · Answer 1 · 2020-09-10

0

Entering edit mode

Ashish Jain ▴ 10

@ashish-jain-14944

Last seen 3.1 years ago

Iowa State University

Hi,

Based on the definition, the genes are defined as "Tissue Enriched" if they have an expression level at least five-fold higher than all the other tissues. This means that for this group a gene is "Tissue Enriched" in only one tissue. Whereas in the case of "Tissue Enhanced" genes, they need to have expression five fold higher than the "average across all other tissues". This means that a gene can be "Tissue Enhanced" in more than one tissue as it can have enhanced expression in more than one tissue. For example "CSF3R" is a tissue enhanced gene in three tissues in the HPA dataset whereas "SLC13A4" is a Tissue Enriched gene only for placenta tissue. You can check this yourself by visiting the web version of TissueEnrich (https://tissueenrich.gdcb.iastate.edu/).

Regards, Ashish Jain

ADD COMMENT • link 4.2 years ago Ashish Jain ▴ 10

0

Entering edit mode

Thanks Ashish Jain!

That's quite clear. But I met another problem with my data. I am not sure if I should open a new issue or not because it looks really wierd! Here is my data. I tried to use teGeneRetrieval only to find that when I set foldChangeThreshold to a high number, I got an error:

example_matrix <- read.table('test.matrix', header = T, row.names = 1, sep = "\t")

se <- SummarizedExperiment(assays = SimpleList(as.matrix(examplematrix)), rowData = row.names(examplematrix), colData = colnames(example_matrix))

output<-teGeneRetrieval(se, foldChangeThreshold = 5) Error in if (sum(idx) > 0) { : missing value where TRUE/FALSE needed

output<-teGeneRetrieval(se, foldChangeThreshold = 3) Error in if (sum(idx) > 0) { : missing value where TRUE/FALSE needed

output<-teGeneRetrieval(se, foldChangeThreshold = 2) Error in if (sum(idx) > 0) { : missing value where TRUE/FALSE needed

However a value 1.5 works fine

output<-teGeneRetrieval(se, foldChangeThreshold = 1.2) output<-teGeneRetrieval(se, foldChangeThreshold = 1.5)

Is this because some of my data is to similar?

ADD REPLY • link 4.2 years ago xiaoxiaoxiaobu • 0

0

Entering edit mode

The problem is the value of "maxNumberOfTissues" argument. In your test data, the tissue number is 5 whereas for calculating group enriched genes you are using 7 (by default). If you change that it to less than or equal to "max(tissues)-1", it will work. This is a good catch and I will put a check for that so that users can know what is going wrong. Usually, we recommend that people use tissueEnrich for data with more than 10 tissues.

ADD REPLY • link 4.2 years ago Ashish Jain ▴ 10

0

Entering edit mode

Many thanks @Ashish Jain ！

Yeah, it will be really nice if a warnning message throws out when the number of input samples are less than "maxNumberOfTissues".

ADD REPLY • link 4.2 years ago xiaoxiaoxiaobu • 0