Question: Meta analysis of RNA-seq data
0
gravatar for bharata1803
3.4 years ago by
bharata1803420
Japan
bharata1803420 wrote:

Hello,

So, currentl, I want to do an anlysis which include 2 spearates experiment from NCBI data sets. The data is generated by Illumina HiSeq 2500 but with a bit different sequencing method. The first is paired end stranded with 48 bp. The second is single end (unknown stranded/unstranded, but I assume stranded because it has the same TruSeq preparation protocol) with 100 bp.

I generated the read count using Salmon. First, I generated human cDNA index from Ensembl GrCH38-80 and do the read counting. I set the parameter of libtype according to the experiment, whether it is single or paired end in Salmon.

Usually, this read count result will only need to be merged and become input for analysis using DESeq2 (if it is rounded to integer) or Limma/voom (as it is).

My question now is, do I need to do some special meta-analysis technique, like published in this paper : http://link.springer.com/article/10.1186%2F1471-2105-15-91

or

because the different in single/paired end is already handled during the read count I can just compare it directly?

Thank you for your answer and suggestion.

rna-seq meta-analysis salmon • 1.6k views
ADD COMMENTlink modified 3.3 years ago by Biostar ♦♦ 20 • written 3.4 years ago by bharata1803420
there are some packages in bioconductor to perform meta-analysis of RNA-seq data. metaSeq is one of these: https://www.bioconductor.org/packages/release/bioc/html/metaSeq.html how does the PCA plot look like? if the study effect is not too big, you can account for it in the linear model.
ADD REPLYlink written 3.4 years ago by Martombo2.4k

Thank you for the suggestion. I will check for the PCA first.

ADD REPLYlink written 3.4 years ago by bharata1803420
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1024 users visited in the last hour