CNVkit: call sex chr - germline analysis

Question

[emma.a]

Hi,

I have 55 males and 45 females and I ran those groups separately and in both cases using a flat reference.

I'm using the following codes

1)
cnvkit.py access ${hg38} -o acces_hg38.bed
cnvkit.py target hg38_design_1-22-X.bed --split --annotate ${flat} -o targets.bed
cnvkit.py antitarget targets.bed -g acces_hg38.bed -o antitargets.bed

2) loop (1 ... N)
cnvkit.py coverage sample1.bam targets.bed -o sample1_targetcoverage.cnn
cnvkit.py coverage sample1.bam antitargets.bed -o sample1_antitargetcoverage.cnn

3)
cnvkit.py reference --sample-sex ${xy} -t targets.bed -a antitargets.bed -f ${hg38} -o FlatReference.cnn

4)
cnvkit.py fix sample1_targetcoverage.cnn sample1_antitargetcoverage.cnn FlatReference.cnn \
-o sample1.cnr
cnvkit.py segment sample1.cnr -o sample1.cns
cnvkit.py segmetrics sample1.cnr -s sample1.cns --ci -o sample1_ci.cns
cnvkit.py call sample1_ci.cns --filter ci --sample-sex ${xy}

When I run the female group for germline CNVs using

--sample-sex x

I do not have any error with sample sex estimation.

On the other side when I run the male group using the option

--sample-sex y

most of the males (not all) are estimated as female. For example

sample.antitargetcoverage.cnn : Male : -1.09 : -0.494
sample.targetcoverage.cnn : Male : -0.711 : -5.96 
sample.cnr : Male : -0.819 : +0.929
sampe.cns : Female : -0.819 : +0.929
sample_ci.call.cns : Female : -0.819 : +0.929 
sample_ci.cns   Female : -0.819 : +0.929

My questions:

1) In general is there any error about steps 1,2,3,4 or any suggestion?

2) Why the pipeline works with the female group and not with the male group? Any suggestion or modification in the code?

Many thanks