GWAS-projections

Cornelis + Andy + Mike 

February 2021
Day 346 since lockdown work from home...

Goal: Prediction of the number of GWAS loci in future neurodegenerative disease GWASes (mainly focussing on the larger ones AD-PD)

Overview of this repo

1) Getting data...

2) Code...

3) Figures...

1) Getting data...

Alzheimer:

Notes: 
- Proxy cases are counted in cases_corrected as 1/2 case
- Of course always the discussion AD vs dementia vs proxy dementia etc. that is ignored here
- 2011 and 2011, two GWAS published but same number of loci so is OK to leave it 
- 2019, two GWAS published a and b added based on chronological order

First_Author	PMID	year	cases	proxies	cases_corrected	controls	loci
Lambert	19734903	2009a	6010	0	6010	8625	3
Harold	19734902	2009b	5964	0	5964	10188	3
Hollingworth	22005930	2011	19870	0	19870	39846	10
Naj	21460841	2011	19490	0	19490	36770	10
Lambert	24162737	2013	25580	0	25580	48466	19
Jansen	30617256	2019a	24087	47793	47983	383378	29
Kunkle	30820047	2019b	35274	0	35274	59163	25
Bellenguez	preprint	2020	64498	46828	87912	677663	75

Parkinson:

Notes: 
- Proxy cases are counted in cases_corrected as 1/2 case
- 2011, two GWAS published a and b added based on chronological order

First_Author	PMID	year	cases	proxies	cases_corrected	controls	loci
Simón-Sánchez	19915575	2009	5074	0	5074	8551	3
Nalls	21292315	2011a	12386	0	12386	21026	11
IPDGC	21738488	2011b	50650	0	15812	21026	16
Nalls	25064009	2014	19061	0	19490	100833	28
Chang	28892059	2017	26035	0	25580	403190	44
Nalls	30617256	2019	37688	18618	46997	1400000	90

2) Code...

library("data.table")
library("ggplot2")
require("ggrepel")

## read in data and predict

data <- read.table("PD_input.txt", header = T)
# OR
data <- read.table("AD_input.txt", header = T)

### prediciton based on cases only
modelCases <- lm(loci ~ cases_corrected, data = data)

### add predictions
data$casePredictedLoci <- predict(modelCases, data)

## look at the data
data

## make the plot for PD
set.seed(42)
basicPlot <- ggplot(data, aes(cases_corrected, loci, label=year)) + geom_point() + geom_smooth(method = "lm", se = T, fullrange = T)
formattedPlot <- basicPlot + theme_bw() + ylab("Number of PD Loci") + xlab("Number of PD Cases") + xlim(0,102000) 
formattedPlotanno <- formattedPlot + geom_text(hjust=-1, vjust=1)
exportPlotPD <- formattedPlot + geom_text_repel(aes(label = data$year), box.padding = 2.5, size = 5, hjust = -1) 

## save plot
ggsave(filename = "lociPlots_PD_new.png", plot = exportPlotPD)

## make the plot for AD
set.seed(42)
basicPlot <- ggplot(data, aes(cases_corrected, loci, label=year)) + geom_point() + geom_smooth(method = "lm", se = T, fullrange = T)
formattedPlot <- basicPlot + theme_bw() + ylab("Number of AD Loci") + xlab("Number of AD Cases") + xlim(0,152000) 
formattedPlotanno <- formattedPlot + geom_text(hjust=-1, vjust=1)
exportPlotAD <- formattedPlot + geom_text_repel(aes(label = data$year), box.padding = 2.5, size = 5, hjust = -1) 

## save plot
ggsave(filename = "lociPlots_AD_new.png", plot = exportPlotAD)

3) Figures...

Alzheimer:

Parkinson:

done...