Hi friends, Do you know how to do a normality test for RNA-seq data set with 60 columns of patients ( 3 groups of phenotypes) and 200 rows of genes?
Also, do you know how to do parametric/non-parametric stat tests for this data? all the documentations on the internet are for one column, but my data has 60 columns.
Thanks
1 answer
There is no point in testing your data for normality, it will fail any test for normality. In general, RNAseq data is not to be normally distributed.
The distribution of counts for one gene across patients is generally assumed to be distributed according the the Negative Binomial Distribution.
The distribution of expressions in one patient across genes (which appears to be what you are talking about as you are talking about your columns), is less well studied, but at least one paper suggests a bimodal mixture of lognormal distributions.
Log in to answer this question.
-
Handling paired before/after microbiome data
written by benkosta •
-
Kruskal Wallis test, gene expression, RNA-seq
written by Rob •
-
RNA-Seq HTseq galaxy DE analysis
written by Rob •
-
proteomics data analysis and non-normality distribution
written by gogeni5529 •
-
copy number variation, fisher exact test, gistic2.0
written by Rob •
-
HT-Seq count data
written by Rob •
-
parallel coordinate plot
written by Rob •
-
correlation matrix for HTSeq data
written by Rob •
-
Normality test for rna seq data
written by Rob •
-
Parametric or non-parametric tests for qPCR significance testing between treatments?
written by n.anuragsharma •
Somewhat duplicate of Normality test for rna seq data