A simple baseline approach based on distance to the known pathogenic variants
MOVA_only_3d_distance.RdThis function evaluates whether pathogenicity can be predicted by the distance between the variant of interest and a known pathogenic variant using a Stratified 5-fold cross validation method.
Usage
MOVA_only_3d_distance(protein_name, MOVA_final_predict_file, MOVA_predict_file, phenotype = "Target")Arguments
- protein_name
Name of target protein/gene. Used to name the file to be exported.
- MOVA_final_predict_file
The "gene name_Target or Pathogenic_finalpredict.csv" file output by the MOVA function. The final MOVA_only_3d_distance predicted values will be in this file.
- MOVA_predict_file
The "gene name_Target or Pathogenic_predict.csv" file output by the MOVA function.
- phenotype
Specify "Target" if you want the positive variant to be the Target variant only, or "Pathogenic" if you want to include the Pathogenic variant as well. The default is "Target".
Details
Draw ROC curve for MOVA_only_3d_distance (black). MOVA_only_3d_distance is evaluated using the Stratified 5-fold cross validation method, and the model is repeated five more times. The average of the distance to the known pathogenic variants is added to the "t_distance" column of the MOVA_predict_file. "protein/gene name_Target or Pathogenic_result_only_3d_distance.csv" contains the Cutoff value (Youden index) for each fold (Column: Cutoff), the number of positive variants for each gene used in the analysis (Column: positive_variant_num), number of negative variants (Column: negative_variant_num), AUC for each fold of MOVA_only_3d_distance (Column: AUC), cvAUC for MOVA_only_3d_distance (Column: cvauc). The file required for redrawing with the MOVA_redraw function of MOVA_only_3d_distance is output in "protein/gene name_Target or Pathogenic_predict_orig_only_3d_distance.csv".
Value
A data.frame containing the Cutoff value (Youden index) for each fold (Column: Cutoff), the number of positive variants for each gene used in the analysis (Column: positive_variant_num), number of negative variants (Column: negative_variant_num), AUC for each fold of MOVA_only_3d_distance (Column: AUC), cvAUC for MOVA_only_3d_distance (Column: cvauc). The same data is output to "protein/gene name_Target or Pathogenic_result_only_3d_distance.csv".
Examples
##---- Should be DIRECTLY executable !! ----
##-- ==> Define data, use random,
##-- or do help(data=index) for the standard data sets.
## The function is currently defined as
function (protein_name, MOVA_final_predict_file, MOVA_predict_file,
phenotype = "Target")
{
D <- fread(MOVA_predict_file)
D$t_distance <- 0
D[, c("t_distance")] <- list(NULL)
if (phenotype == "Target") {
D <- D[(D$Type == "Target") | (D$Type == "Ctrl"), ]
}
fpredict <- data.frame(matrix(rep(NA, 2), nrow = 1))[numeric(0),
]
colnames(fpredict) <- c("ID", "distance")
final_data <- fread(MOVA_final_predict_file)
df <- data.frame(matrix(rep(NA, 6), nrow = 1))[numeric(0),
]
colnames(df) <- c("ID", "re_result", "distance", "change",
"iter1", "iter2")
D$id2 <- 0
D[, c("id2")] <- list(NULL)
D$re_result <- 0
D[, c("re_result")] <- list(NULL)
D$t_distance <- 0
D[, c("t_distance")] <- list(NULL)
D$iter1 <- 0
D[, c("iter1")] <- list(NULL)
D$iter2 <- 0
D[, c("iter2")] <- list(NULL)
if (phenotype == "Target") {
D$result <- D$Type2
}
else {
D$result <- D$Type3
}
oldw <- getOption("warn")
options(warn = -1)
D <- rowid_to_column(D, var = "id2")
for (i2 in 1:5) {
D$id3 <- D$id2
j <- 5
dat1 <- D %>% stratified(., group = "result", size = 1/j)
dat1$iter <- 1
datzan <- D[-dat1$id2, ]
for (i in 2:j) {
datzan[, c("id2")] <- list(NULL)
datzan <- rowid_to_column(datzan, var = "id2")
if (i != j) {
dat2 <- datzan %>% stratified(., group = "result",
size = 1/(j - i + 1))
}
else {
dat2 <- datzan
}
dat2$iter <- i
datzan <- datzan[-dat2$id2, ]
dat1 <- rbind(dat1, dat2)
}
dat1[, c("id2")] <- list(NULL)
dat1 <- rowid_to_column(dat1, var = "id2")
for (i in 1:j) {
val <- dat1[dat1$iter == i, ]
train.data <- dat1[-val$id2, ]
x <- train.data[train.data$result == 1, ]$x
y <- train.data[train.data$result == 1, ]$y
z <- train.data[train.data$result == 1, ]$z
for (i3 in 1:nrow(val)) {
f <- data.frame(ID = val[i3]$ID, re_result = abs(val[i3]$result -
1), t_distance = min(sqrt((x - val[i3]$x) *
(x - val[i3]$x) + (y - val[i3]$y) * (y - val[i3]$y) +
(z - val[i3]$z) * (z - val[i3]$z))), change = val[i3]$change,
iter1 = i, iter2 = i2)
df <- rbind(df, f)
}
}
}
options(warn = oldw)
df$iter3 <- df$iter1 + (df$iter2 * 5)
out <- cvAUC(df$t_distance, df$re_result, label.ordering = NULL,
folds = df$iter3)
plot(out$perf, col = "black", avg = "vertical", add = TRUE)
cvauc <- out$cvAUC
YI2 <- data.frame(matrix(rep(NA, 5), nrow = 1))[numeric(0),
]
colnames(YI2) <- c("Cutoff", "positive_variant_num", "negative_variant_num",
"AUC", "cvauc")
for (i in 6:30) {
pred <- prediction(df[df$iter3 == i, ]$t_distance, df[df$iter3 ==
i, ]$re_result)
auc.tmp <- performance(pred, "auc")
auc <- as.numeric(auc.tmp@y.values)
tab <- data.frame(Cutoff = unlist(pred@cutoffs), TP = unlist(pred@tp),
FP = unlist(pred@fp), FN = unlist(pred@fn), TN = unlist(pred@tn),
Sensitivity = unlist(pred@tp)/(unlist(pred@tp) +
unlist(pred@fn)), Specificity = unlist(pred@tn)/(unlist(pred@fp) +
unlist(pred@tn)), Accuracy = ((unlist(pred@tp) +
unlist(pred@tn))/nrow(df)), Precision = (unlist(pred@tp)/(unlist(pred@tp) +
unlist(pred@fp))))
tab$Youden <- tab$Sensitivity + tab$Specificity - 1
YI <- tab[order(tab$Youden, decreasing = T), ]
YI <- data.frame(Cutoff = YI$Cutoff[1], positive_variant_num = c(nrow(df[(df$re_result ==
0) & (df$iter2 == 1), ])), negative_variant_num = c(nrow(df[(df$re_result ==
1) & (df$iter2 == 1), ])), AUC = c(auc), cvauc = c(cvauc))
YI2 <- rbind(YI2, YI)
}
fwrite(YI2, paste(protein_name, phenotype, "result_only_3d_distance.csv",
sep = "_"))
df2 <- df %>% group_by(ID) %>% summarise(t_distance = mean(t_distance))
D[, c("result")] <- list(NULL)
fwrite(merge(D, df2), MOVA_predict_file)
fwrite(df, paste(protein_name, phenotype, "predict_orig_only_3d_distance.csv",
sep = "_"))
return(YI2)
}
#> function (protein_name, MOVA_final_predict_file, MOVA_predict_file,
#> phenotype = "Target")
#> {
#> D <- fread(MOVA_predict_file)
#> D$t_distance <- 0
#> D[, c("t_distance")] <- list(NULL)
#> if (phenotype == "Target") {
#> D <- D[(D$Type == "Target") | (D$Type == "Ctrl"), ]
#> }
#> fpredict <- data.frame(matrix(rep(NA, 2), nrow = 1))[numeric(0),
#> ]
#> colnames(fpredict) <- c("ID", "distance")
#> final_data <- fread(MOVA_final_predict_file)
#> df <- data.frame(matrix(rep(NA, 6), nrow = 1))[numeric(0),
#> ]
#> colnames(df) <- c("ID", "re_result", "distance", "change",
#> "iter1", "iter2")
#> D$id2 <- 0
#> D[, c("id2")] <- list(NULL)
#> D$re_result <- 0
#> D[, c("re_result")] <- list(NULL)
#> D$t_distance <- 0
#> D[, c("t_distance")] <- list(NULL)
#> D$iter1 <- 0
#> D[, c("iter1")] <- list(NULL)
#> D$iter2 <- 0
#> D[, c("iter2")] <- list(NULL)
#> if (phenotype == "Target") {
#> D$result <- D$Type2
#> }
#> else {
#> D$result <- D$Type3
#> }
#> oldw <- getOption("warn")
#> options(warn = -1)
#> D <- rowid_to_column(D, var = "id2")
#> for (i2 in 1:5) {
#> D$id3 <- D$id2
#> j <- 5
#> dat1 <- D %>% stratified(., group = "result", size = 1/j)
#> dat1$iter <- 1
#> datzan <- D[-dat1$id2, ]
#> for (i in 2:j) {
#> datzan[, c("id2")] <- list(NULL)
#> datzan <- rowid_to_column(datzan, var = "id2")
#> if (i != j) {
#> dat2 <- datzan %>% stratified(., group = "result",
#> size = 1/(j - i + 1))
#> }
#> else {
#> dat2 <- datzan
#> }
#> dat2$iter <- i
#> datzan <- datzan[-dat2$id2, ]
#> dat1 <- rbind(dat1, dat2)
#> }
#> dat1[, c("id2")] <- list(NULL)
#> dat1 <- rowid_to_column(dat1, var = "id2")
#> for (i in 1:j) {
#> val <- dat1[dat1$iter == i, ]
#> train.data <- dat1[-val$id2, ]
#> x <- train.data[train.data$result == 1, ]$x
#> y <- train.data[train.data$result == 1, ]$y
#> z <- train.data[train.data$result == 1, ]$z
#> for (i3 in 1:nrow(val)) {
#> f <- data.frame(ID = val[i3]$ID, re_result = abs(val[i3]$result -
#> 1), t_distance = min(sqrt((x - val[i3]$x) *
#> (x - val[i3]$x) + (y - val[i3]$y) * (y - val[i3]$y) +
#> (z - val[i3]$z) * (z - val[i3]$z))), change = val[i3]$change,
#> iter1 = i, iter2 = i2)
#> df <- rbind(df, f)
#> }
#> }
#> }
#> options(warn = oldw)
#> df$iter3 <- df$iter1 + (df$iter2 * 5)
#> out <- cvAUC(df$t_distance, df$re_result, label.ordering = NULL,
#> folds = df$iter3)
#> plot(out$perf, col = "black", avg = "vertical", add = TRUE)
#> cvauc <- out$cvAUC
#> YI2 <- data.frame(matrix(rep(NA, 5), nrow = 1))[numeric(0),
#> ]
#> colnames(YI2) <- c("Cutoff", "positive_variant_num", "negative_variant_num",
#> "AUC", "cvauc")
#> for (i in 6:30) {
#> pred <- prediction(df[df$iter3 == i, ]$t_distance, df[df$iter3 ==
#> i, ]$re_result)
#> auc.tmp <- performance(pred, "auc")
#> auc <- as.numeric(auc.tmp@y.values)
#> tab <- data.frame(Cutoff = unlist(pred@cutoffs), TP = unlist(pred@tp),
#> FP = unlist(pred@fp), FN = unlist(pred@fn), TN = unlist(pred@tn),
#> Sensitivity = unlist(pred@tp)/(unlist(pred@tp) +
#> unlist(pred@fn)), Specificity = unlist(pred@tn)/(unlist(pred@fp) +
#> unlist(pred@tn)), Accuracy = ((unlist(pred@tp) +
#> unlist(pred@tn))/nrow(df)), Precision = (unlist(pred@tp)/(unlist(pred@tp) +
#> unlist(pred@fp))))
#> tab$Youden <- tab$Sensitivity + tab$Specificity - 1
#> YI <- tab[order(tab$Youden, decreasing = T), ]
#> YI <- data.frame(Cutoff = YI$Cutoff[1], positive_variant_num = c(nrow(df[(df$re_result ==
#> 0) & (df$iter2 == 1), ])), negative_variant_num = c(nrow(df[(df$re_result ==
#> 1) & (df$iter2 == 1), ])), AUC = c(auc), cvauc = c(cvauc))
#> YI2 <- rbind(YI2, YI)
#> }
#> fwrite(YI2, paste(protein_name, phenotype, "result_only_3d_distance.csv",
#> sep = "_"))
#> df2 <- df %>% group_by(ID) %>% summarise(t_distance = mean(t_distance))
#> D[, c("result")] <- list(NULL)
#> fwrite(merge(D, df2), MOVA_predict_file)
#> fwrite(df, paste(protein_name, phenotype, "predict_orig_only_3d_distance.csv",
#> sep = "_"))
#> return(YI2)
#> }
#> <environment: 0x0000011a8737a180>