Tissue Specific Markers in Extracellular Vesicles

This repository contains code which identifies tissue specific markers based on the GTEx and roadmap epigenomics project databases, as well as additional data from SRA and GEO.

The primary question that is being asked (and hopefully answered) is:

1. How can we identify the source tissue of exosomes found in a peripheral fluid? 2. What are genes whose expression is particular to a tissue? 3. Knowing those genes, can we identify the source tissue of in silico generated exosome reads?

Contents of the Repo

1. common_r contains various R utilities for sub analyses
2. exosome_reads contains code to download placenta and various other tissue SRA experiments, perform an alignment, bias and subset the aligned reads to exosomes, and then call the resultant subset+biased bam files using cufflinks, and then pile all of the data up. See its Makefile for details.
3. exosome_sra contains code to examine reads from various extracellular vesicles from different tissue types. Currently not used for anything.
4. go_analyses contains the code to perform GO analyses on the basis of the identified tissue specific markers. See its Makefile for details.
5. isrs_abstract is an abstract for the Illinois Symposium on Reproductive Science which was accepted
6. manuscript contains the code to produce the paper based upon this work. See its Makefile for details.
7. mk contains various makefile settings
8. poster contains the code for a poster which was presented at the Mayo Individualizing Medicine conference.
9. rnaseq_workflow is a git submodule which is the standard RNAseq workflow
10. slides is a selection of slides for the ISRS talk
11. tissue_expression contains all of the code to build an R matrix of tissue expression across a large number of samples. See its Makefile for details.
12. tissue_specific_expression contains the code to turn the matrix of tissue expression into a list of genes which are possibly tissue specific and genes which are not by calculating τ. It also contains code to run SVM and other routines using caret to do machine learning to identify source tissues.