Can I use TCGA-LUAD RNAseq count that had already normalized by RSEM in Limma-voom
1
0
Entering edit mode
3 months ago
Fluke ▴ 10

Hi everyone, first of all, I'm new for bioinformatics. I have downloaded RNAseq data of TCGA-LUAD from UCSC that had already normalized RSEM normalized count and log2 transformed (log2 normcount+1). i wonder if i can use this dataset to do differential expressed genes using limma-voom?

RNAseq differential-gene-expession limma batch-effect • 691 views
ADD COMMENT
0
Entering edit mode

Yes, normalized RSEM counts from TCGA can be used as input for Limma Voom. Please check this post https://support.bioconductor.org/p/63981/#64004

ADD REPLY
0
Entering edit mode

Thanks a lot for your answer, I’m confused between RSEM expected count and RSEM normalized count. I’m working on this dataset from UCSC https://xenabrowser.net/datapages/?dataset=TCGA.LUAD.sampleMap%2FHiSeqV2&host=https%3A%2F%2Ftcga.xenahubs.net&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443 It’s “.rsem.genes.normalized_results” and as i understand, they just do batch effect normalization of RSEM values. Could i use it instead of RSEM expected count?

ADD REPLY
0
Entering edit mode

After reads are mapped to the genome, there are algorithms that quantify the number of mapped reads onto a genomic locus. Some of these algorithms only count incidences where a read is mapped to a single genomic locus, whereas other algorithms may include multi mapping reads to their quantification. Expected counts refer to gene expression values quantified via the latter approach, since gene expression has to be estimated when accounting for multi mapping reads.

Normalised counts meant that the raw counts were normalised.

ADD REPLY
0
Entering edit mode
12 weeks ago
jv ★ 1.6k

There is no mention within Xena that the data has been batch corrected, nor is the term "normalized" commonly used to refer to batch corrected data. I think this post nicely describes what the RSEM "normalized" counts are.

For my own work with RSEM counts I always start with the expected counts and treat as "raw" counts data. With expected counts you can do any number of processing steps to prepare you data for downstream analyses and there will be no ambiguity about what has been done to the data b/c you will have done it your self.

ADD COMMENT

Login before adding your answer.

Traffic: 1557 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6