Similar trends from RNA-Seq values

[priya]

Hi,

I have time-series RNA-Seq data . Both FPKM and counts values are available. Especially for the DE analysis, Counts are transformed to voomed values and all the DE-tests are performed. I would like to find out the group of genes that are having similar trend over time.

Eg, I would like to pick the genes that are showing similar trend like GeneA

Gene A :
T0 T1 T2 T3 T4 T5
10 20 30 25 10 2

Is there any proper approach to categorize the genes .
Any method or papers or suggestions will of great help.

[mattanswers]

supervised clustering

I would suggest supervised clustering. You can input a given gene's expression pattern and the rest of the genes will be clustered by what genes show the most similar patterns.

Class Neighbors and/or Gene Neighbors in GenePattern, http://http://www.broadinstitute.org...re/genepattern, will do what you want.

There is also STEM,http://http://www.cs.cmu.edu/~jernst/stem/, and some publications that you can get by googling 'gene expression time series'.

[priya]

I would suggest supervised clustering. You can input a given gene's expression pattern and the rest of the genes will be clustered by what genes show the most similar patterns.

Class Neighbors and/or Gene Neighbors in GenePattern, http://http://www.broadinstitute.org...re/genepattern, will do what you want.

There is also STEM,http://http://www.cs.cmu.edu/~jernst/stem/, and some publications that you can get by googling 'gene expression time series'.

Thank you for your suggestions and links. It is exactly what I was looking for,