Abnormal chromosomal X z-scores

[AnnickDaniels]

Hi all,

I've been comparing two versions of WisecondorX: version 1.1.5 vs version 1.2.7, since we want to update our pipeline. Comparison is mainly based on chromosomal z-scores (from (chr_)statistics.txt) of more than 200 samples analyzed with both pipelines. It seems z-scores correlate very well for each chromosome. However, the average z-score of chromosome X is -4.8 in version 1.2.7 versus -0.6 in version 1.1.5 ??!!

Additional information: each pipeline has its own reference comprised of more than 500 normal samples (about 50%/50% F/M). Different samples were taken for each reference. Binsize for each was 50kb. --nipt parameter was used for reference generation.
The reference genome for both was GRCh38, however different versions. To create the bowtie index for our latest pipeline
I downloaded the genome reference from
https://ftp.ensembl.org/pub/release-112/fasta/homo_sapiens/dna/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz.

Any ideas on what is going on and what to do about it are very welcome.