VarTrix calls reference reads in a cell line that does not contained reference reads

[dcruz5h]

Hello ! First thank you so much for generating this pipeline as it is really easy to use.
I run VarTrix on a 10X run from a cell mix experiment. I mixed three cell lines at equal ratios:
My gene of interest falls in the X-chromosome if the cell line is male there is only one allele and if its female there are two alleles.

1. THP1-A cell line that does not express my gene of interest but it is wild type at this locus (male)
2. K562-A cell line that expressed my gene of interest and it is wild type at this locus (female)
3. CMK-A cell line that expressed my gene of interest but is mutant at this locus (male)

As you can see from the plot below, while THP1 cells cannot be called which make sense as they don't express my gene of interest, and K562 cells are for the majority WT, some cells cannot be called which is expected as this depends on the level of expression of my gene of interest.

CMK cells which should all be mutant, about 20% of them have at least one reference read.

Would mind speculating Where are these reference reads coming from in CMK cells.
is there cross contamination during cell capture in the 10X platform ?

vlnplot1.pdf