integrate new tests

Merge branch 'rework-tests' into predict_gain

jass/test/data_test1/COV.csv

-ID	z_DISNEY_RATATOUY	z_DISNEY_POCAHONT
-z_DISNEY_RATATOUY	2.05403006060606	0.394332909090909
-z_DISNEY_POCAHONT	0.394332909090909	1.17729254545455

jass/test/data_test1/SumStatTab.txt

-rsid	position	chr	region	Z_ratatouy	Z_pocahont
-rs1	1	1	1	0.812	-1.06
-rs2	2	1	1	2.197	0.937
-rs3	3	1	2	2.049	0.854
-rs4	1	2	3	1.632	1.461
-rs5	2	2	3	0.254	-1.413
-rs6	3	2	4	0.491	0.567
-rs7	4	2	4	-0.324	0.583
-rs8	5	2	4	-1.662	-1.307
-rs9	1	3	5	1.768	-0.54
-rs10	2	3	5	0.026	1.948
-rs11	3	3	6	1.129	0.054
-rs12	4	3	7	-2.38	0.352

jass/test/data_test1/metadata.txt

-information	content
-title	Mock dataset with disney
-description	"lorem ipsum"
-ancestry	DIS
-assembly	dSNY

jass/test/data_test1/regions.txt

-chr	start	stop
-chr1	1	2
-chr1	3	4
-chr2	1	2
-chr2	3	5
-chr3	1	2
-chr3	3	3
-chr3	4	4

jass/test/data_test1/summary.csv

-Outcome	FullName	Consortium	Type	Reference	ReferenceLink	dataLink	internalDataLink	Nsample	Ncase	Ncontrol
-RATATOUY	Ratatouille ou la mort de l'hygiène en cuisine	DISNEY	BrainWashing	Courgette et al., 1754	http://www.marmiton.org/recettes/recette_ratatouille_23223.asp	pouet	pouet	1000000		
-POCAHONT	Pocahontas mange des tapas	DISNEY	BrainWashing	Rolfe et al., 1614	https://fr.wikipedia.org/wiki/Pocahontas	Gargar	Gargar	1000000		

jass/test/data_test1/z_DISNEY_POCAHONT_chr1.txt

-rsid	position	refAllele	altAllele	Z
-rs1	1	C	T	-1.06
-rs2	2	G	T	0.937
-rs3	3	C	T	0.854

jass/test/data_test1/z_DISNEY_POCAHONT_chr2.txt

-rsid	position	refAllele	altAllele	Z
-rs4	1	G	T	1.461
-rs5	2	A	T	-1.413
-rs6	3	A	C	0.567
-rs7	4	A	G	0.583
-rs8	5	C	G	-1.307

jass/test/data_test1/z_DISNEY_POCAHONT_chr3.txt

-rsid	position	refAllele	altAllele	Z
-rs9	1	C	T	-0.54
-rs10	2	G	T	1.948
-rs11	3	A	C	0.054
-rs12	4	A	C	0.352

jass/test/data_test1/z_DISNEY_RATATOUY_chr1.txt

-rsid	position	refAllele	altAllele	Z
-rs1	1	C	T	0.812
-rs2	2	G	T	2.197
-rs3	3	C	T	2.049

jass/test/data_test1/z_DISNEY_RATATOUY_chr2.txt

-rsid	position	refAllele	altAllele	Z
-rs4	1	G	T	1.632
-rs5	2	A	T	0.254
-rs6	3	A	C	0.491
-rs7	4	A	G	-0.324
-rs8	5	C	G	-1.662

jass/test/data_test1/z_DISNEY_RATATOUY_chr3.txt

-rsid	position	refAllele	altAllele	Z
-rs9	1	C	T	1.768
-rs10	2	G	T	0.026
-rs11	3	A	C	1.129
-rs12	4	A	C	-2.38

jass/test/data_test2/COV.csv

-ID	z_BMW_ISETTA	z_BMW_MINI	z_FIAT_CINQCENT	z_FIAT_CINQUECENTO	z_MERCO_SMART	z_TATA_TATANANO
-z_BMW_ISETTA	2	1	1	1	1	1
-z_BMW_MINI	1	2	1	1	1	1
-z_FIAT_CINQCENT	1	1	2	1	1	1
-z_FIAT_CINQUECENTO	1	1	1	2	1	1
-z_MERCO_SMART	1	1	1	1	2	1
-z_TATA_TATANANO	1	1	1	1	1	2

jass/test/data_test2/create_unit_test_smart.R

-## There 6 phenotypes
-sumtab <- read.table("summary.txt", sep="\t", header=TRUE, stringsAsFactors = FALSE)
-## Zscore ID: z_CONSORITUM_PHENOTYPE_chr#chr.txt
-ids <- sprintf("z_%s_%s",sumtab$Consortium, sumtab$Outcome)
-## The covariance is set to 1 and the variance to 2
-COV <- toeplitz(c(2,1,1,1,1,1))
-rownames(COV) <- colnames(COV) <- ids
-
-
-## Structure:
-# - 5 chromosomes, 2 regions per chromosomes
-# - 10 regions, 2 regions per chromosome
-# - 30 SNPs, 3 SNPs per region
-
-## Structure of missing values region per region ;
-##   ".." means no missing values
-##   "XX" means the whole region is missing
-#
-#      Z1 Z2 Z3 Z4 Z5 Z6
-# R1   .. .. .. .. .. ..
-# R2   .. .. .. .. .. ..
-# R3   XX .. .. .. .. ..
-# R4   .. XX .. .. .. .. 
-# R5   .. .. XX .. .. ..
-# R6   .. .. .. XX .. ..
-# R7   .. .. .. .. XX ..
-# R8   .. .. .. .. .. XX
-# R9   XX XX XX .. .. .. 
-# R10  .. .. .. XX XX XX
-filenames <- paste0(rep(ids,e=5), "_chr", rep(1:5, 6), ".txt")
-
-## rsid : rs_#chr_#region_#snp
-rsid <- paste0("rs", "_chr", rep(1:5, e=6), # chr
-               "_reg", sprintf("%02i", rep(1:10, e=3)), # region
-               "_", sprintf("%02i", 1:30)) # snp
-pos <- rep(1:6, 5)
-chr <- rep(1:5, e=6)
-reg <- rep(1:10, e=3)
-ref <- "A"
-alt <- "G"
-BIGZ <- matrix(1:(30*6), 30, 6)
-BIGZ[grep("reg03", rsid),1] <- NA
-BIGZ[grep("reg04", rsid),2] <- NA
-BIGZ[grep("reg05", rsid),3] <- NA
-BIGZ[grep("reg06", rsid),4] <- NA
-BIGZ[grep("reg07", rsid),5] <- NA
-BIGZ[grep("reg08", rsid),6] <- NA
-BIGZ[grep("reg09", rsid),1:3] <- NA
-BIGZ[grep("reg10", rsid),4:6] <- NA
-rownames(BIGZ) <- rsid
-
-# What does it look like ?
-require(pheatmap)
-png("zscores.png", res = 100)
-pheatmap(BIGZ, cluster_rows = FALSE, cluster_cols = FALSE, cellwidth = 10, cellheight = 10)
-dev.off()
-# Create regions
-
-# Create Z scores
-
-
-# Write covariance matrix
-write.table(data.frame(ID=ids, COV), file="COV.csv", row.names = F, quote=F, sep="\t")
-
-# Write region file
-regions <- data.frame(chr=sprintf("chr%i", rep(1:5,e=2)),
-                      start=rep(c(1,4),5),
-                      stop=rep(c(3,6),5))
-write.table(regions, file="regions.txt", row.names = F, quote=F, sep="\t")
-# Write all the Z files
-k <- 1
-for (j in 1:6) {
-  for (chrnum in 1:5) {
-    ind <- grep(sprintf("chr%i", chrnum), rsid)
-    tmp <- data.frame(rsid = rsid[ind], 
-                      pos = pos[ind],
-                      ref=ref,
-                      alt=alt,
-                      Zscore=unname(BIGZ[ind, j]))
-    write.table(na.omit(tmp), file = filenames[k], quote=FALSE)
-    k <- k+1
-  }
-}
-
-# Compute the summary statistic
-
-