Thanks in advance for even read this.
I am new in bioinformatic (3-4 month with rudimentary R programming knowledge). I need some constructive comment before I embark on this stuff. I am writing my PhD thesis (due in 6 month). The part that I write below should be the 1st half of my thesis (the 2nd half will be metabolites data from multiple gene edited mice).
Let say, I have 12 genes of interest ("A - M") which might be related to certain diseases or "X" biological pathway. Our Department already have "X" pathway metabolites data from gene edited mice (KO/KD) from all these 12 genes.
My main/general objective is to identify which of these 12 genes is the most important a) amongst themselves and b) in system biology in general.
To answer this I thought Meta-Analyses of Gene Expression Data might be a good idea to answer a by comparing the 12 genes expression pattern and in various associated disease model to answer b.
Here is the main Question How to explain and study pattern?. Right now, my plan is, using PRISMA Systematic Review guide to filter Mice MicroArray Dataset (because our metabolites data is from mice and I only kinda adept using R to manipulate MicroArray dataset), and then conduct Meta-Analyses of the datasets using P-value combination, Fisher's Method.
But then after that I am lost. What should I do next?
Thank you for you insight!
I am still doen't know how big are this idea. I have been wandering what to include in the first part of my dissertation (English is not my first language, and ive been using thesis and dissertation interchangebly). No this is not PhD in US.
The main reason for this idea is for me to tap into vast metabolites data that the department have with bioinformatic gene studies.
For the past few month I have read some studies and prepare R code where I can download and transform my test datasets from GEO to be load into Network Analyst webtools, or independently provide DEG using limma and provide output for me to load into DAVID (for GO) and from DAVID output prepare a table to be paired with STRINGApp nodes in Cytoscape for a nice clustering in Cytoscape (kinda like ClusterGO, just discovered later it exist, but I can control what GO to be cluster with my method). Not sure whether any of this is useful, but it have been a fun and exciting learning experience (procrastinating from writing?)
I am focusing on MicroArray as quick GEO database skimming I find lots of studies model with my pathway of interest. So I ask, why not try do meta-analyses on that dataset and see result on my gene of interest. Now I am preparing simple presentation slide to propose this to my supervisor, if he says no, then that's the end.
Anyway thank you Dr. Mensur for your advice. My mentor also discourage me, although in my thought I just think because he is not well verse in this field that is why he's not interested. I just wish I had more time to do this or discover this stuff much sooner.