inst/dataprep/H56-CD4.R

In MiguelRodo/VaccCompData: Data package for novel TBV vaccine comparison paper

# Loading -----------------------------------------------------------------

# names of current objects
currObjVec = ls()

# find library containing raw data
pkgPath = system.file( package = "VaccCompData" )

# libraries
source( file = paste0( pkgPath, "/dataprep/libraries.R") )

#read in data
fileLoc = paste0( pkgPath, "/extdata/h56.xlsx" )
cd4Tbl = as_tibble( read.xlsx( fileLoc, sheet = 'CD4 Freq') )

# Data Prep ---------------------------------------------------------------

#UNIQUE VALUES

#unique values in id info
unique(cd4Tbl$Stimulation) # "Ag85B"  "ESAT6"  "PHA"    "Rv2660" "UNS"
unique(cd4Tbl$Study.Day) #0  14  56  70 292 112 126. Correct
unique(cd4Tbl$LTBI) # "neg" "pos". Correct, but convert to 0 1
unique(cd4Tbl$Group) # 1 2 3 4. Correct
unique(cd4Tbl$`Vaccine/Placebo`) # "vaccine" "placebo". Correct, but convert to 0 1
unique(cd4Tbl$Dose) # 15 5 50 0 "?". Correct
unique(cd4Tbl$`Regimen.(#injections)`) # 2 1 3. Corrrect.

# renaming columns and converting to correct type
cd4Tbl %<>% mutate( Stimulation = str_to_lower(Stimulation) ) %>%
  rename( group = Group, inf = LTBI, ptid = PTID, stim = Stimulation, timePoint = Study.Day, treatType = `Vaccine/Placebo`, dose = Dose ) %>%
  mutate( group = as.character( group ), inf = as.character( inf ), dose = as.character( dose ), timePoint = as.character( timePoint ) ) %>%
  rename( infxn = inf )

# cd4Tbl %>%
 # filter( ptid == 'ID_03013' & timePoint == 70 ) # simply missing. Checked in raw data - not there either.

# dimension names
dimnames(cd4Tbl)[[2]][8] = 'doseCount'

for( i in 9:29 ) cd4Tbl[[ i ]] = as.numeric( cd4Tbl[[ i ]] )

# change doseCount variable type
cd4Tbl %<>% mutate( doseCount = as.character( doseCount ) )

# fill in ?s

# find the locations
#cd4Tbl %>%
#  filter( treatType == "?" | dose == "?") %>%
#  print( n = 10^5 )

# finding replacement values

# cd8Tbl %>% filter( ptid == 'id_01015' & timePoint == 292 ) # dose = 50 & treatType = 1
# cd8Tbl %>% filter( ptid == 'id_02003' & timePoint == 0 ) # dose = 5 and treatType = 1
# cd8Tbl %>% filter( ptid == 'id_02012' & timePoint == 0 ) # dose = 5 & treatType = 1

# replacement
#cd4Tbl[ cd4Tbl$ptid == 'ID_01015', ]$dose = 50
#cd4Tbl[ cd4Tbl$ptid == 'ID_01015', ]$treatType = 'vaccine'
cd4Tbl[ cd4Tbl$ptid == 'h56_15', ]$dose = 50
cd4Tbl[ cd4Tbl$ptid == 'h56_15', ]$treatType = 'vaccine'

#cd4Tbl[ cd4Tbl$ptid == 'ID_02003', ]$dose = 5
#cd4Tbl[ cd4Tbl$ptid == 'ID_02003', ]$treatType = 'vaccine'
cd4Tbl[ cd4Tbl$ptid == 'h56_53', ]$dose = 5
cd4Tbl[ cd4Tbl$ptid == 'h56_53', ]$treatType = 'vaccine'


#cd4Tbl[ cd4Tbl$ptid == 'ID_02012', ]$dose = 5
#cd4Tbl[ cd4Tbl$ptid == 'ID_02012', ]$treatType = 'vaccine'
cd4Tbl[ cd4Tbl$ptid == 'h56_62', ]$dose = 5
cd4Tbl[ cd4Tbl$ptid == 'h56_62', ]$treatType = 'vaccine'

# editing entries
cd4Tbl %<>%
  mutate( infxn = sapply( infxn, function(x) ifelse( x == 'neg', '0', '1' ) ),
          ptid = str_to_lower( ptid ),
          stim = str_to_lower( stim ),
          treatType = sapply( treatType, function(x) ifelse( x == 'vaccine', '1', '0' ) ) )

### BOOLEAN COLUMN NAMES

# remove useless columns
cd4Tbl %<>% select( -c(`CD4+`, `g-T-2-17-`, `CD4+g+`, `CD4+T+`, `CD4+2+`, `CD4+17+` ))

# get actual names
cytNameVec = str_to_upper( tbl_vars(cd4Tbl)[9:23] )

# get replacement names
for (i in 1:length(cytNameVec)){ #loop over names
  currCyt = cytNameVec[i]
  #check if we have a positive in a given location
  #if so, add a p onto a string. If not, add an n
  locVec = c(2,4,6,9)
  signVec = str_sub(currCyt, locVec,locVec)
  for (j in 1:length(signVec)){
    if (signVec[j] == '+'){
      str_sub(currCyt, locVec[j], locVec[j]) = 'p'
    } else if (signVec[j] == '-'){
      str_sub(currCyt, locVec[j], locVec[j]) = 'n'
    }
  }
  cytNameVec[i] = currCyt
}

# replace names

origCytNameVec = tbl_vars(cd4Tbl)[9:23]
colNameCodeDF = data_frame(x = str_c("cd4Tbl %>% rename( cd4", cytNameVec, " = `", origCytNameVec, "`)"))

cd4Tbl = eval(parse(text = colNameCodeDF[1,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[2,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[3,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[4,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[5,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[6,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[7,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[8,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[9,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[10,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[11,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[12,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[13,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[14,1]))
cd4Tbl = eval(parse(text = colNameCodeDF[15,1]))

### remove placebo
cd4Tbl %<>%
  filter( treatType == 1 ) %>%
  select( -treatType )

### RE-ORDER

cd4Tbl = changeCytOrder( dataTibble = cd4Tbl, firstCytColIndex = 8,
                          firstCytNameLoc = 4, signLocVec = c( 5, 7, 9, 12 ),
                          orderLocVec = c( 1, 3, 2, 4 ), cdPresent = TRUE )

# Data Checking -----------------------------------------------------------

# stim
# four stims
#cd4Tbl %>%
#  group_by( inf, group, ptid, timePoint ) %>%
#  summarise( stimCount = n() ) %>%
#  filter( stimCount != 5 ) # id_03008 & id_03019, both on day 14

#viewTbl = cd4Tbl %>%
#  filter( ptid == 'id_03008' | ptid == 'id_03019' ) %>%
#  filter( timePoint == 14 )
viewTbl = cd4Tbl %>%
  filter( ptid == ' h56_86' | ptid == 'h56_90' ) %>%
  filter( timePoint == 14 ) %>%
  select( group:doseCount ) # both are missing unstim.

cd4Tbl %>%
  arrange( infxn, group, ptid, timePoint, stim ) %>%
  group_by( infxn, group, ptid, timePoint ) %>%
  summarise( stimVec = str_c(stim, sep = "", collapse = "" ) ) %>%
  filter( stimVec != "ag85besat6pharv2660uns") # same as above, so everyone else has all the right stims for each day.

# cd4Tbl %<>%
#  filter( !( ptid == 'id_03008' & timePoint == '14' ) ) %>%
#  filter( !( ptid == 'id_03019' & timePoint == '14' ) )

cd4Tbl %<>%
  filter( !( ptid == 'h56_86' & timePoint == '14' ) ) %>%
  filter( !( ptid == 'h56_90' & timePoint == '14' ) )


# Subtraction ---------------------------------------------------------------

# save non-summed over table
cd4ITbl = cd4Tbl %>%
  group_by( group, infxn, ptid, dose, doseCount, timePoint ) %>%
  mutate_if( is.numeric, function(x) x - x[5] ) %>%
  ungroup() %>%
  filter( stim %in% c( "ag85b", "esat6", "rv2660" ) )

# summation
cd4Tbl %<>%
  group_by( group, infxn, ptid, dose, doseCount, timePoint ) %>%
  summarise_if( is.numeric, function(x) sum( x[ c( 1, 2, 4 ) ] ) - 3 * x[ 5 ] ) %>%
  ungroup()
# unique( cd4Tbl %>% filter( group == ) ) %>% length()

# Editing -----------------------------------------------------------------

h564ITbl = cd4ITbl %>%
  mutate( timePoint = as.numeric( timePoint ) )

h564Tbl = cd4Tbl %>%
  mutate( timePoint = as.numeric( timePoint ) )

# Testing -----------------------------------------------------------------

# h564ITbl
testRow1 = h564ITbl %>% filter( ptid == 'h56_1' & timePoint == 0 & stim == "ag85b" )
if( testRow1$CD4Gn2nTn17p != -0.001679 ) stop( "h564ITbl error")
testRow2 = h564ITbl %>% filter( ptid == 'h56_32' & timePoint == 292 & stim == "esat6" )
if( testRow2$CD4Gp2nTp17n != ( 0.000815 - 0.000465 ) ) stop( "h564ITbl error" )
testRow3 = h564ITbl %>% filter( ptid == 'h56_97' & timePoint == 292 & stim == "esat6" )
if( testRow3$CD4Gp2nTn17n != ( 0.030077 - 0.003363 ) ) stop( "h564ITbl error" )
# h564Tbl
testRow1 = h564Tbl %>% filter( ptid == 'h56_1' & timePoint == 0 )
if( signif( testRow1$CD4Gn2nTp17n, 6 ) != signif( ( 0.009479 + 0.125443 + 0.010263 - 3 * 0.008241 ), 7 ) ) stop( "h564Tbl error")
if( testRow1$CD4Gn2pTn17n != ( 0.002916 + 0.000725 + 0.002932 - 3 * 0.000634 ) ) stop( "h564Tbl error")
if( testRow1$CD4Gp2pTp17n != ( 0.013853 + 0.008701 + 0 - 3 * 0 ) ) stop( "h564Tbl error" )
# Clear Workspace ---------------------------------------------------------

# remove all objects but initial objects (currObjVec) and processed data tibble (h564Tbl)
rm( list = setdiff( ls(), c( currObjVec, "h564Tbl", "h564ITbl" ) ) )