KAR Protocol

  1. Creating Input Files
    1. Kinase Profile
    2. The kinase profile file is a 'binarized' table of the effect of each drug on a large panel of kinases. Values are either 0 (inhibited) or 100 (not inhibited). Example files for H1581 and H1975 are included with the algorithm download. The Python version uses CSV files and the MATLAB version uses Microsoft Excel files.

      Example:
      Drug Name Kinase 1 Kinase 2 Kinase 3 Kinase 4 Kinase 5
      A-443654 0 100 100 100 100
      A-770041 100 0 0 100 100
      AS-601245 0 100 100 100 100
      AZ-628 100 0 0 100 100
      AZD-7762 0 100 100 0 0
      BI-2536 100 100 100 100 100
      BMS-536924 0 100 0 0 100
      Bosutinib 0 0 100 100 100
      CGP-60474 100 100 100 100 100

    3. Drug Sensitivity
    4. The drug sensitivity file is a table containing the results of a drug screen. Column 1 contains the drug names and column 2 contains the sensitivity readout (e.g. IC50, Percent of control) for each drug. The order and number of drugs should be identical to the kinase profile table.

      Example:
      Drug Name H1975
      A-443654 0.20
      A-770041 0.29
      AS-601245 1.33
      AZ-628 9.48
      AZD-7762 1.17
      BI-2536 0.11
      BMS-536924 0.34
      Bosutinib 1.39

    5. Gene Expression
    6. The last input file is a table of gene expression values for each kinase. The order and number of kinases must be identical to the kinase profile table. Column 1 contains a list of kinases and column 2 contains gene expression measurements for each kinase. Here, we used publicly available normalized microarray data, but the gene expression threshold could be modified to accommodate other data types such as RNA-seq.

      Example:
      AAK1 9.54
      ABL1 8.83
      ABL2 5.04
      ACVR1 9.13
      ACVR1B 9.24
      ACVR2A 7.66
      ACVR2B 6.61
      ACVRL1 4.81
      ADCK3 8.47
      ADCK4 6.01
      AKT1 8.91
      AKT2 7.27
      AKT3 6.71

  2. Running KAR
    1. Python version(KAR.py)
    2. Open terminal and enter the following commands:
      >> python
      >> import KAR
      >> KAR.KAR(var1, var2, var3, var4, var5, var6, var7, var8, var9, var10)
      >> exit()

      Variables 1-10 are defined as follows:
      • Var 1: output filename for single kinase scores (e.g. 'single_output.txt')
      • Var 2 : output filename for scores for kinase pairs (e.g. 'combo_output.txt')
      • Var 3 : CSV filename drug kinase profile data (e.g. 'Kinase_profile.csv').
      • Var 4 : CSV filename with drug sensitivity data (e.g. 'drug_sensitivity.csv')
      • Var 5 : CSV filename with gene expression data (e.g. 'gene_ex.csv')
      • Var 6 : Threshold for gene expression (kinases with lower gene expression values will be filtered, default =7)
      • Var 7 : Threshold for highest sensitivity bin - drugs with a sensitivity value less than threshold with be in Bin 1 (default=1)
      • Var 8 : Threshold for high sensitivity bin - drugs with a sensitivity value greater than or equal to Var7 and less than Var 8 will be in Bin 2 (default =2).
      • Var 9 : Intermediate sensitivity threshold - drugs with a sensitivity value greater than or equal to Var8 and less than Var9 will be in Bin 3 (default =5)
      • Var10 : Highly insensitive threshold - drugs with a sensitivity values greater than or equal to Var9 and less than Var10 will be in Bin 4. Drugs with a sensitivity value greater than or equal to Var10 will be in negative scoring Bin 5 (default=10).

      Example:

      >> KAR.KAR('H1975_single.txt','H1975_combo.txt','H1975_Kinase Binary Profiles.csv','H1975_Drug Sensitivity.csv','H1975_Gene Expression.csv',7,1,2,5,10)

    3. MATLAB version

    4. Open and run MATLAB file KAR.m.

      1. Fill in desired output filenames when prompted by the pop-up window.
      2. Select Excel files with drug kinase profiles, drug sensitivity data, and gene expression data when prompted.
      3. Enter the following thresholds into the pop-up window when prompted:
        1. Gene expression threshold (kinases with lower gene expression values will be filtered, e.g. =7)
        2. Threshold for highest sensitivity bin - drugs with a sensitivity value less than threshold with be in Bin 1 (e.g.=1)
        3. Threshold for high sensitivity bin - drugs with a sensitivity value greater than or equal to Var7 and less than Var8 will be in Bin 2 (e.g.=2).
        4. Intermediate sensitivity threshold - drugs with a sensitivity value greater than or equal to Var8 and less than Var9 will be in Bin 3 (default = 5)
        5. Highly insensitive threshold - drugs with a sensitivity values greater than or equal to Var9 and less than Var10 will be in Bin 4. Drugs with a sensitivity value greater than or equal to Var10 will be in negative scoring Bin 5 (default=10).

KAR for TNBC