Project Description

• The purpose of this project is to create an algorithm/model that is able to take datasets generated from similar studies and combine them.
• The goal in combining the data from the different data sets in such a way as to reduce the noise that comes from each being a different data set but retain the biological signal present in the data.
• Similar problems have been tackled by many others, but this specific problem has not been approached before as we hope to.
• We want to create an algorithm that is as generalizable as possible.
• The data used for this process can be found on OSF at this link
  • This data was collected and cleaned by Ifeanyichukwu Nwosu, a former PhD student of Dr. Piccolo
• I have attempted to keep track of all of the computation that I have done in the form of R Scripts. Each of these scripts should be run from within the root directory of the project, not from within the "scripts/" folder.
  • At the top of many of the scripts is a setwd() function that sets the working directory to my specific root directory on my computer. This line should be changed to reflect the location of your root directory in order to work properly.
• To get a better idea of what the eventual model I want to build looks like, see my Model Plan
• Also, a note about the links in my Updates:
  • They work with Pandoc when I make the MD files into PDF files, but apparently not in GitHub, sorry about that.

Repository Description

• This repository holds all of the information about the work that I am doing as part of the Piccolo Lab in the Life Sciences Department of Brigham Young University
• This repository contains several folders, as outlined below, though it does not contain all of the folders that I have in my working directory as I work on this project.

Working Directory Contents

• These are the folders that I have in my working directory:
  • data/
    • This folder holds the raw gene expression data obtained from OSF
    • The following data sets can be found at this link rather than on OSF
      • GSE123845
      • GSE115577
      • GSE163882
  • functions/
    • This folder holds a set of functions that I have written to help in several tasks.
  • merged/
    • This folder holds all of the data sets after the metadata and gene expression data has been merged by the collectMerged() function
  • meta/
    • This folder holds all of the metadata for the different data sets obtained from OSF
    • The following data sets can be found at this link rather than on OSF:
      • GSE123845
      • GSE115577
      • GSE163882
  • models/
    • This folder holds all of the fit keras models that I use in my analysis, so I don't have to recreate them when I need to use them again.
  • others/
    • This folder holds the other random things I use in generating documents and testing things in my project.
  • plots/
    • This folder holds all of the plots I have generated during my analysis for various reasons
  • pdfs/
    • This folder contains pdf versions of my updates
  • scripts/
    • This folder holds all of the scripts that I use to run my analysis
  • updates/
    • These are Markdown files in which I keep track of the work I am doing, be it taking notes on ideas or the workflow/process of what I am accomplishing.
  • variables/
    • This folder holds any variables that are particularly important in my analysis that I don't want to have to obtain again when I need them.
• Some of these folders are not present in this repository because they only contain data, which I do not want to take up room in the repository since it is already stored in OSF
• Furthermore, my models are too large to store on GitHub, so they are not found here.
• Hopefully that helps in explaining what everything is for and in finding anything of interest

Other Files

• A list of my completed and yet unfinished tasks is found in my TODO file