Cutoff for TPM value
Hi,
I need to filter my transcriptome to speed up some process.
I will map my samples against my transcriptome and compute the TPM value per contig per sample.
Do you have an idea about which minimum TPM value I should use if any sample has it ?
• 5,874 views
•
link
updated
by
Matteo Schiavinato
•
written
by
Picasa
•
written
by
Picasa
1 answer
One of the "established" procedures is the one of taking FPKM=1, and then converting it to TPM by doing:
TPM_threshold = [ 1 / sum(FPKM_sample) ] * 10^6
This would give you a sample-specific TPM threshold which is scaled to the FPKM=1 threshold, used in many publications. However, you should keep in mind that also FPKM=1 was merely a choice.
• 0 views
•
link
written
by
Matteo Schiavinato
Log in to answer this question.
More posts like this
-
Normalization strategies for comparing mapped reads across samples in RNA-seq
written by Picasa
-
What is the recommended contig length to filter?
written by DNAngel
-
Get counts for allele specific expression analysis
written by Picasa
-
Trying to find best way to determine co-occurence between taxa and genes related to antibiotic resi…
written by Longshotx •
-
Filter transcripts based on expression value: TPM vs TMM ?
written by Picasa
-
How to compute global chromosome expression from RNA-Seq
written by Nicolas Rosewick
-
SSPACE option (-k and -a) for scaffolding
written by Picasa
-
Confusion about the kmer coverage
written by Picasa
-
Blast a transcriptome to genome ?
written by Picasa
-
Annotating A Novel Transcriptome Assembly
written by Madelaine Gogol
Repeat of this question?
Your question is answered here