searchDEE2.Rmd

---
title: "Get a list of human RNA-seq studies"
author: "The Gene Sig Commons Group"
date: "08/07/2020"
output: html_document
---

## Intro

In this script we're obtaining the current list of human RNA-seq studies from dee2.io using the getDEE2 R package.

```{r cars}
library("getDEE2")
md <- getDEE2::getDEE2Metadata("hsapiens")
head(md)
dim(md)
length(unique(md$GEO_series))
head(md$GEO_series)
dee2_gse <- as.character(unique(md$GEO_series))
length(dee2_gse)
```

## Extract GDS information

Keywords: Epilepsy, heart disease/CVD, diabetes.
Go to https://www.ncbi.nlm.nih.gov/gds/ and search your disease of interest. 
Refine the search results for Homo sapiens only.
Save the results as a file; "send to" --> "file" --> "summary"
Use Rstudio to upload the file and then the below code will work if the file names match.

```{r,importgds}
geo <- readLines("gds_result.txt")
head(geo,10)
geo <- geo[grep("GSE",geo)]
geo <- geo[grep("Accession",geo)]
geo <- sapply(strsplit(geo," "),"[[",2)
geo <- gsub("\tID:","",geo)
geo <- unique(geo)
head(geo)
length(geo)
```

## Intersect

With this part you will get a list of studies (GEO series) that are relatd to your disease and present in the DEE2 database. 

```{r,itx}
intersect(dee2_gse,geo)
```

I need you to make a new text file in the GitHub repo, for example "epilepsy.md" which describes each of the intersected studies and whether they are good candidates for processing by us:

* The study is relevant to the disease.

* The experiment is replicated, that means n>2