If you work on multimorbidity, network medicine, sex/gender in health, or drug repurposing, we’ve made these results (and more) available in our portal: disease-perception.bsc.es/shdc/
@alfonsovalencia.bsky.social @lmrocha.bsky.social
doi.org/10.1038/s438...

One striking case: metformin, used to treat type 2 diabetes, appears linked to a reduced risk of liver cancer, with a stronger protective signal in men than in women.

Another example: smoking → COPD.
In women, we observe overactivation of cytochrome P450 enzymes and immune pathways.
In men, we instead see downregulation of DNA repair processes.
We also asked whether we could connect drugs to these sex-specific relationships. Again, the answer is yes.

We then asked: Can two diseases co-occur in both sexes but via different biological processes? The answer is clearly yes.
Example: type 2 diabetes & pancreatic cancer.
In women, antiviral responses and regulation of cholesterol biosynthesis.
In men, changes in extracellular matrix organisation.

Digging into the underlying biology, we found examples where the higher COPD risk in women with schizophrenia, or the higher IBD risk in women who smoke, may be driven by immune, metabolic, and ROS-detoxification pathways.

In other words, the sex-stratified networks explain about 53–60% of comorbidities that differ between women and men, giving us molecular hypotheses for why some multimorbidity patterns are sex-specific.

When we built the networks, we saw that, similar to epidemiology, 13–16% of transcriptomic similarities are sex-specific. These sex-specific links molecularly connect disease pairs that co-occur more than expected by chance in only one sex.

In this work, we collected ~9k samples across 100+ diseases and constructed separate disease similarity networks for women and men to examine how molecular profiles align with epidemiological comorbidity.
​

Pleased to finally share this: our paper on sex-specific transcriptomic similarity networks to study comorbidities is published! This started from a simple but annoying gap: we know that women and men have different multimorbidity patterns, but we still rarely look at the underlying biology by sex.

Before lunch break Elma Dervic gave us a very quick talk on how to build comorbidity networks from data. I loved her visualization through time. More amazing talks at #NetBioMed2025 after lunch, so stay tuned! #NetSci2025 @verapancaldi.bsky.social