required data:

* train.RData
* Phenotable_AML_Array_Cohorte_488_patients_03_2017_Readout.xlsx
* probes.Rdata
* Phenotable_AML_Array_Cohorte_488_ELN_Var.xlsx
* test.Rdata
* Pheno_Test_Full.csv

######################

# training_data_preparation.R		
  preprocessing the training data

# model_calc_restricted.R		
  calculation of models with restrictions to the number of coefficients
  requires data from training_data_preparation.R

# model_calc_unrestricted.R		
  calculation of models without restrictions to the number of coefficients
  requires data from training_data_preparation.R



######################
###   validation   ###
######################


# the following scripts are based on model_calc_restricted.R. For model_calc_unrestricted.R, it works analogously 


# Scores_Train.R
  calculating the scores for training data
  requires data from training_data_preparation.R and model_calc_restricted.R
 

# Scores_Test.R
  calculating the scores for test data
  requires data from model_calc_restricted.R


# Cutoffs.R
  calculating the cutoffs for TPR, TNR, and AUC
  requires data from training_data_preparation.R and Scores_Train.R


# Groups.R
  Produce three risk groups for training and test data
  requires data from training_data_preparation.R, cutoffs.R, Scores_Train.R, Scores_Test.R


# Table3.R
  first half of Table 3: TPR, TNR, AUC, C-index, IBS. calculation of prediction error curves.
  requires data from Groups.R, Scores_Train.R, model_calc_restricted.R

# Optimism.R
  calculating optimism for Table 3
  requires data from training_data_preparation.R, Scores_Train.R, Scores_Test.R


# KM_Curves.R
  Kaplan-Meier-Plots for Discrimination
  requires data from training_data_preparation.R, Groups.R


# HR_pval.R
  Second half of Table 3: Hazard ratios across risk groups, CIs, and p-values of the LR-Test
  requires data from Groups.R


# Calibration.R
  Kaplan-Meier-Plots for Calibration
  requires data from training_data_preparation.R, Groups.R
