statistic test
Hello, I am naive to the statistics. for example, MACS2 detects peaks using Poisson distribution, but why does it can output p value? Another example ia RNAseq, some algorithms use NB distribution to model gene expression, and then use Wald test perform differential analysis. Why not directly using t test or other statistic test?
• 1,002 views
•
link
updated
by
Istvan Albert
•
written
by
BenHu •
•
written
by
BenHu •
1 answer
The distribution type is meant to capture the expected characteristics of that data.
A p-value reflects the likelihood of the null hypothesis being true for the chosen distribution.
Think of a test as a formula that accounts for various features of the data.
If you google the concepts you will find an extremely wide array of explanations for each.
• 0 views
•
link
written
by
Istvan Albert
Log in to answer this question.
More posts like this
-
Is a log2 transformation an essential step in preparing expression data for machine learning?
written by Sib •
-
the distribution of log2Fc and t value are not similar for bulk RNA-seq
written by alwayshope •
-
Why does my t-test fail?
written by C_sinensis •
-
Can I use All Peaks (MACS2 -q 1) for analysis?
written by Tian •
-
RNA seq analysis; two samples
written by Lpak •
-
meta analysis vs Cross Platform normalization
written by jmgoldstein7 •
-
Simulation of a 1.25-fold change using Poisson Data
written by nkhan.mscs15seecs •
-
Using LRT and Wald test to filter differentially expressed loci in RNA-seq dataset
written by a.rex
-
Algorithmic Difference between MACS and MACS2
written by taraeicher •
-
Negative Binomial and Poisson distribution of RNA-Seq
written by ashwini
