In MyersGroup/testisAtlas: testisAtlas: R functions and analysis scripts used in the analysis of testis single cell RNAseq

old_names <- readRDS("../data/count_matrices/V3_SDAmerged_mouse_V3_SDA_dimnames.rds")

# backup old dimnames
saveRDS(old_names, "../data/count_matrices/V3_SDAmerged_mouse_V3_SDA_dimnames_archive.rds")

# get ENSEMBL ID for old names
ensembl <- useMart("ENSEMBL_MART_ENSEMBL", host = "http://aug2017.archive.ensembl.org", dataset = "mmusculus_gene_ensembl")
mapTab <- getBM(attributes = c("external_gene_name",'ensembl_gene_id'),
                filter = "external_gene_name", values = old_names[[2]], mart = ensembl, uniqueRows=TRUE)
mapTab <- data.table(mapTab)
mapTab[external_gene_name %in% mapTab[duplicated(external_gene_name)]$external_gene_name]

# get new names from IDs
ensembl <- useMart("ENSEMBL_MART_ENSEMBL", host = "http://Dec2017.archive.ensembl.org", dataset = "mmusculus_gene_ensembl")
mapTab2 <- data.table(getBM(attributes = c("external_gene_name",'ensembl_gene_id'),
                filter = "ensembl_gene_id", values = mapTab$ensembl_gene_id, mart = ensembl, uniqueRows=TRUE))

setkey(mapTab, ensembl_gene_id)
setkey(mapTab2, ensembl_gene_id)

mapTabM <- merge(mapTab, mapTab2, all.x=T)
mapTabM <- mapTabM[!duplicated(mapTabM)]

# for genes with duplicated external_gene_name.x, are the external_gene_name.y different anyway?
mapTabM[external_gene_name.x %in% mapTabM[duplicated(external_gene_name.x)]$external_gene_name.x][order(external_gene_name.x)]
mapTabM[external_gene_name.x %in% mapTabM[duplicated(external_gene_name.x)]$external_gene_name.x][external_gene_name.x!=external_gene_name.y]


tmp <- mapTabM[match(old_names[[2]], external_gene_name.x),]
# which gene names updated
tmp[external_gene_name.y!=external_gene_name.x]

# which gene names NA (deleted in new database), use old IDs as names
tmp[is.na(external_gene_name.y)]
tmp[is.na(external_gene_name.y), external_gene_name.y := ensembl_gene_id]

# which gene names duplicated, use old IDs as names
tmp[external_gene_name.y %in% tmp[duplicated(external_gene_name.y)]$external_gene_name.y][order(external_gene_name.y)]
tmp[external_gene_name.y %in% tmp[duplicated(external_gene_name.y)]$external_gene_name.y & external_gene_name.y!=external_gene_name.x ,external_gene_name.y := ensembl_gene_id]

tmp[is.na(external_gene_name.y) | duplicated(external_gene_name.y)]

str(tmp$external_gene_name.y)
str(old_names[[2]])

tail(tmp$external_gene_name.y)
tail(old_names[[2]])

old_names[[2]] <- tmp$external_gene_name.y
saveRDS(old_names, "../data/count_matrices/V3_SDAmerged_mouse_V3_SDA_dimnames.rds")
saveRDS(tmp, "../data/count_matrices/V3_SDAmerged_mouse_V3_SDA_dimnames_ID_mapping.rds")

library(biomaRt)
ensembl <- useMart("ENSEMBL_MART_ENSEMBL", host = "http://Dec2017.archive.ensembl.org", dataset = "mmusculus_gene_ensembl")
mapTab1 <- data.table(getBM(attributes = c("external_gene_name",'ensembl_gene_id'),mart = ensembl, uniqueRows=TRUE))

saveRDS(mapTab1, "../data/Ensembl_92_gene_symbols.rds")

MyersGroup/testisAtlas documentation built on Nov. 29, 2020, 9:21 p.m.