Dear All,

I'm trying to identify driver events from CNV data using oncodriveCIS. It's essentially requires a use of Copy number and expression data.

OncodriveCIS is provided with GBM example data and this one works perfectly.

However when I'm doing the analysis with my own data

python oncodrivecis.py \
  -e $DATA_PATH/expression.sampleswithSNP.arrays.tsv \
  -c $DATA_PATH/CNV.SNP.array.TARGET.NB.and.normals.tsv \
  -s $DATA_PATH/samples.to.process.SNP.arrays.tsv \
  -o $DATA_PATH/20200723_OncodriveCIS_SNParrays \
  -i $DATA_PATH/ensembl63_ensembl2hugo.tsv -p NB

I'm getting this type of an error:

File "oncodrivecis.py", line 59, in oncodriveCIS
    amp_eisN_dict, amp_eisT_dict = calculate_eis(options, cnv_dict, '2')   File "/gpfs01/home/x/oncodriveCIS/oncodrivecis-1.1.0/src/lib.py", line 171, in calculate_eis
    eisN_subdict = get_eis(exp_header_s, normal_positions, tumor_alt_positions, exp_l_s)   File "/gpfs01/home/x/oncodriveCIS/oncodrivecis-1.1.0/src/lib.py", line 69, in get_eis
    eis = eis_calculation(float(exp_l_s[tumor_alt_position]), ref_median, ref_IQ, tumor_alt_IQ)   File "/gpfs01/home/x/oncodriveCIS/oncodrivecis-1.1.0/src/lib.py", line 52, in eis_calculation
    eis = (exp_v - ref_median) / (abs(ref_IQ) + abs(tumor_alt_IQ)) ZeroDivisionError: float division by zero

I have checked and I don't have any zeroes in my dataset. I also eliminated genes with low expression levels. Any ideas what I could improve?