How to compare gene expression datasets from 3 different GPLs.
Hi guys, I hace microarray data GSE data from 4 different GPLs. I have also downloaded the series expression matrix for all the GSE files and also the annotation file for all the GPLs that i am using.My doubt is how to combine these 4 series expression data.Do I have to normalise them?Or just on basis of gene ID I can combine.
• 170 views
•
link
0 answers
No answers yet.
Log in to answer this question.
You probably cannot put different platforms into the same statistical analysis. There are all kinds of platform-specific batch effects such as exact sequence and number of probes per gene that would skew the analysis. Better analyze them independently and perform meta-analysis on them.
Is there any way to take care of the batch effect problem?And also if I analyse them individually how to metaanalyse them for WGCNA
I would start by extensively reviewing literature, tools and posts about batch correction and WGCNA and then come back with more specific questions. I do not think people feel like giving step-by-step protocols and code for your question given that you do not seem to have invested any effort yourself (at least you do not point out what you've tried in this and our earlier posts on the same matter). Others have already asked you in these earlier posts to show some effort and to read the manuals of the tools you aim to use. Check
limma's batch correction function and inform yourself on what batches are, and what the requirements are to correct them.Converting multiple GSE data into expression matrix How to enter gene list in R for WGCNA analysis
SIr I have actually worked on tha lst advice and created the combined dataset.But the batch effect is something new to me.
Hi @ATpoint I actually did go through the batch effect and WGCNA literature.And from there I got to know about various methods of merging.So I wanna know if inSilicomerging and Insilico db seem to be a good option in merging my files.kindly help.
Hey dude / dudette, can you please help us to try to keep the forum neat by not creating multiple questions that are ultimately about the same thing? Here is your new question: Which is the best r package for co expression analysis