Question: heatmap for coding and noncoding genes
0
gravatar for rhasanvandj
18 days ago by
rhasanvandj10
rhasanvandj10 wrote:

Hi friends, I did differential expression for rna seq data: I did for coding genes also I did for all genes(coding+noncoding). I want to show significant genes in heatmap. Do you think heatmap is good for significant coding genes or all genes?

rna-seq • 125 views
ADD COMMENTlink modified 18 days ago by cristopher.biotec10 • written 18 days ago by rhasanvandj10
4
gravatar for Kevin Blighe
18 days ago by
Kevin Blighe66k
Kevin Blighe66k wrote:

It is your choice and it depends on the message that you want to convey in your work. You do not really need anybody else's verification for this task.

  • All genes == unsupervised clustering
  • Statistically significant genes == supervised clustering

Kevin

ADD COMMENTlink modified 18 days ago • written 18 days ago by Kevin Blighe66k

I may have missed your point slightly in relation to non-coding versus protein coding genes. If you have conducted your entire analysis separately for protein coding genes and non-coding genes (that is, normalisation and differential expression analysis separately), then it is my view that they also should be shown separately in different heatmaps.

There is no problem conducting your analysis by just focusing on protein coding genes alone, and there is neither a problem focusing on protein coding + non-coding combined.

As you can see, there are no rules here... you can present your results in the way that helps to get the information best to the reader.

ADD REPLYlink written 18 days ago by Kevin Blighe66k

Thanks Kevin I have differentially expressed genes for coding genes alone and also all genes together(Coding+noncoding). I have volcano plot and now I want to plot heatmap for them.

ADD REPLYlink written 18 days ago by rhasanvandj10

For the heatmap, you could try to follow this: A simple tutorial for a complex ComplexHeatmap

, or, at the most basic level, you just need:

heatmap(assay(vst)[sigGenes,])

, assuming here that you have used DESeq2 and that your normalsied+transformed data is contained in assay(vst) and that your vector of statistically significant genes, whose ID type should match that of the rownames of assay(vst), are held in sigGenes.

ADD REPLYlink modified 18 days ago • written 18 days ago by Kevin Blighe66k
1
gravatar for cristopher.biotec
18 days ago by
cristopher.biotec10 wrote:

You can show those results in the enhanced volcano plot using a different shape for coding and non-coding genes

ADD COMMENTlink written 18 days ago by cristopher.biotec10

Good suggestion - I am also a developer of EnhancedVolcano.

ADD REPLYlink written 18 days ago by Kevin Blighe66k

Hi Cristopher Thanks good suggestion. I did volcano plot before just for significant genes all same shape. But how can I do this with different shape for coding and non-coding genes?

ADD REPLYlink written 15 days ago by rhasanvandj10
1

Please see the vignette, which I personally wrote: https://bioconductor.org/packages/devel/bioc/vignettes/EnhancedVolcano/inst/doc/EnhancedVolcano.html#over-ride-colour-andor-shape-scheme-with-custom-key-value-pairs

ADD REPLYlink written 15 days ago by Kevin Blighe66k
1

Thanks Kevin I use your EnhancedVolcano a lot. This is really amazing. Appreciate that.

ADD REPLYlink written 15 days ago by rhasanvandj10

Thank you Sir. You got it working?

ADD REPLYlink written 15 days ago by Kevin Blighe66k

Yes dear Kevin Your Enhancedvolcano is awesome and very useful. I wish you would do something like this for heatmap.

thanks

ADD REPLYlink written 12 days ago by rhasanvandj10
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1944 users visited in the last hour