Quick MR of fibronectin / cytokines / matrix metalloproteinases on osteoarthritis

mr_inflammation_oa.rmd

---
title: Quick MR of fibronectin / cytokines / matrix metalloproteinases on osteoarthritis
date: "`r Sys.time()"
---


This paper presents a mathematical model linking osteoarthritis (OA) to a dynamic inflammation system involving cytokines and fibronectin https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5562782/.

Quick look at whether genetic relationships associate:

Trait IDs:

```{r}
library(TwoSampleMR)
library(MRInstruments)
library(tidyverse)

ao <- available_outcomes()

oa_id <- "ebi-a-GCST007092"
fibronectin_id <- c("prot-a-1131", "prot-a-1132", "prot-a-1133", "prot-c-4131_72_2")
cytokine_id <- c("prot-a-683", "prot-a-684", "prot-a-1520", "prot-a-666", "prot-a-667", "prot-a-668", "prot-a-2355")
crp_id <- "ieu-b-35"
il6_id <- "prot-a-1539"
il1_id <- ao %>%
	filter(grepl("Interleukin-1 ", trait)) %>%
	{.$id}
il10_id <- ao %>%
	filter(grepl("Interleukin-10 ", trait)) %>%
	{.$id}
tnfalpha_id <- "prot-c-3722_49_2"
mmp_id <- ao %>%
	filter(grepl("Matrix metallopr", trait)) %>%
	{.$id}
```

Perform MR

```{r}
dat <- make_dat(c(fibronectin_id, cytokine_id, il1_id, il6_id, crp_id, il10_id, tnfalpha_id, mmp_id), oa_id)
o <- mr(dat, method_list=c("mr_ivw", "mr_wald_ratio"))
o %>% 
	filter(method %in% c("Inverse variance weighted", "Wald ratio")) %>%
	ggplot(., aes(x=b, y=exposure)) +
	geom_point() +
	geom_errorbarh(aes(xmin=b-1.96*se, xmax=b+1.96*se), height=0) +
	geom_vline(xintercept=0, linetype="dotted")
```

No obvious associations here. Wrong cell-type? Perhaps look at gene-expression levels in the correct cell-type instead? Could use 

```{r}
data(gtex_eqtl)
table(gtex_eqtl$tissue)
```

Are any of these relevant tissue types?