#!/usr/bin/env Rscript library (randomForest) library (MASS) ################ # MAIN # ################ args <- commandArgs(TRUE) pdesc2D = args[1] pdescOPERA = args[2] pmodel = args[3] pout = args[4] #pdesc2D = "./interferences/temp/376251/2D.csv" #pdescOPERA = "./interferences/temp/376251/OPERA.csv" #pmodel = "./interferences/static/interferences/models/AUTOFLUORESCENCE_BLUE/model7.Rdata" #pout = "./interferences/temp/376251/temp.txt" load(pmodel) model = outmodel$model #print (model) # open 2D din2D = read.csv(pdesc2D, sep = "\t", header = TRUE) # remove duplicate name din2D = din2D[-which(duplicated(din2D[,1])),] rownames(din2D) = din2D[,1] # del SMILES colunm if("SMILES" %in% colnames(din2D)){ din2D = din2D[,-which(colnames(din2D) == "SMILES")] } if("inchikey" %in% colnames(din2D)){ din2D = din2D[,-which(colnames(din2D) == "inchikey")] } #open OPERA dinOPERA = read.csv(pdescOPERA, sep = "\t", header = TRUE) dinOPERA = dinOPERA[-which(duplicated(dinOPERA[,1])),] rownames(dinOPERA) = dinOPERA[,1] lchem = intersect(rownames(din2D), rownames(dinOPERA)) ddesc = cbind(din2D[lchem,], dinOPERA[lchem,]) #ddesc = ddesc[,-1] ############################### check error in the interpred modeling lpred = predict(model, ddesc) dout = cbind(lchem, lpred) colnames(dout) = c("ID", "pred") rownames(dout) = rownames(dout) write.csv(dout, pout)