13/13 We are excited to finally be able to share these results with the community, and would love to get your feedback and thoughts!

12/13 Relatedly, Xi genes may be under unusually strong selection. @ipsitaagarwal.bsky.social‬ and colleagues showed that the fitness cost of heterozygote loss of function for a group of NPX Xi genes with Y homologs is higher than any other matched group of genes. elifesciences.org/articles/103...

11/13 HbA1c, a marker of glucose control, may reflect Xi expression effects in metabolic tissues like the liver or pancreas—where genes like GYG2 and PNPLA4 exhibit high Xi expression.
Schizophrenia is a disorder with known sex differences and previous observations of XCI skew in affected females.

10/13 We hypothesize that this mechanism may contribute to the correlation of Xi expression and dominance signals for lymphocyte %, HbA1c, and schizophrenia.
A heritable component to X chromosome inactivation skew may help explain the greater incidence of autoimmune disorders in females.

9/13 As a concrete example, one genome-wide significant dominance effect on HbA1c levels is in the body of the ZFP92 gene, which encodes for zinc finger protein that regulates DNA transcription and transposable element silencing and may plausibly influence X chromosome inactivation.

8/13 In trying to make sense of dominance signals and Xi specifically, we formalized a simple expectation for dominance effects that is unique to X-linked variants: a variant will produce a dominance effect on a trait if it is also linked to a QTL for X chromosome inactivation.

7/13 The significance of additive allelic effects grows with Xi expression levels. Critically, dominance signals (when alleles bias which X is inactivated) also track Xi escape in several traits, including lymphocyte %, HbA1c, and schizophrenia.

6/13 We then asked: do these Xi‐escape levels predict where X‐linked genetic variation matters most in women? Using UK Biobank X‐chromosome GWAS on diseases and physiological traits, we tested both additive and dominance models.

5/13 Even genes not formally called “escapees” fall on this line—escape is a spectrum. This suggests that other compensatory or regulatory mechanisms play a more minor role in sex differences in X-linked gene expression

4/13 We were taken back by how predictive Xi expression levels were for sex differences and variation within females. In particular, Xi expression was a strong linear predictor of female–male expression difference across non‐pseudoautosomal (NPX) genes.

3/13 We started by quantifying escape as a continuous trait (proportion of expression from Xi) using allele‐specific data from a single GTEx individual with “skewed” X-inactivation, where virtually all cells in all tissues have the X inherited from the same parent inactivated.

2/13 In XX females one X is largely silenced, but ~⅕ of X‐linked genes “escape” silencing to varying degrees. We were fascinated by work from Adrianna San Roman and colleagues showing that expression from Xi itself has major effects on gene expression genomewide. www.cell.com/cell-genomic...

1/13 In a new preprint, we (with @xliaoyi.bsky.social‬ in the ‪@arbelharpak.bsky.social‬ lab) find that expression from the “inactivated” X (Xi) is consequential for both female-male differences in gene expression and variation among females in disease and physiology: www.biorxiv.org/content/10.1...