FAQ      Help   
  
Normalization of imported data
Question
ID
141
Category
Qlucore Omics Explorer
  Data Import
Date Created
2021-05-05 11:43:25
Date Updated
2021-05-05 11:46:00
RNASeq analysis sometimes produces e.g. .tsv files with several different normalizations. Which normalization should I use for analysis in Qlucore?

* Raw Counts: Raw (integer) counts per gene and sample.

* Normalised Counts: Normalised (real) counts per gene and sample scaled
to the sequencing depth by DESeq2.

* VST Blind Counts: Counts on a log2 scale per gene and sample after
variance stabilising transformation (VST) disregarding the generalised
linear model design.

* VST Model Counts: Counts on a log2 scale per gene and sample after
variance stabilising transformation (VST) considering the generalised
linear model design.

* FPKMs: Fragments Per Kilobase Per Million.


Answer
Use

* Raw Counts if you want to use the normalizations available i Qlucore (FPKM/TPM/TMM).

* VST Blind Counts if you want to use the VST normalization from DESeq2.

* VST Model Counts if you know which model was used with DESeq2 and want to use that. An alternative approach is to use VST Blind Counts normalization + eliminated factors in Qlucore.

* FPKM requires log transform of data. It is preferable to instead use the raw counts + the FPKM in Qlucore since the software then will handle the log transformation.

Related articles
Was this information helpful?
 
Back to Search Results