I was running this tool called ContEst ( https://www.broadinstitute.org/cancer/cga/contest_run ) by Broad Institute to determine NGS sample contamination. I ran the analysis twice and got two different results for the same samples. The contamination level I am getting for the first analysis is less than 0.5% and in the second analysis, it's more than 97% for the both samples. Can someone please explain the reason for these variegated results?

First Analysis results:

AOGC_NAME   population  population_fit  contamination   confidence_interval_95_width    confidence_interval_95_low  confidence_interval_95_high sites
S09-F09-P01     CEU n/a 0.3 0.1 0.3 0.4 4498
AOGC-02-0112    CEU n/a 0.5 0.1 0.5 0.6 4870

Second Analysis results:

AOGC_NAME   population  population_fit  contamination   confidence_interval_95_width    confidence_interval_95_low  confidence_interval_95_high sites
S09-F09-P01     META    CEU n/a 97.8    0.4 97.6    98  4208
AOGC-02-0112    META    CEU n/a 97.7    0.6 97.4    98  4619