Collab with @shaicarmi.bsky.social - Paper here:
www.nature.com/articles/s41...

Conclusion: PES is quite distinct from classical PGT-M for high-penetrance variants. However, PGT-M for moderate penetrance variants has recently been adopted by some clinics. This practice is more like PES, necessitating careful consideration of all factors affecting risk.

3) Through modeling and simulation, we show that PGT-M for a moderate penetrance variant would select the "wrong" (i.e., higher risk) embryo ~5% of the time (given at least 1 carrier and 1non-carrier embryo, both female). IOW, the non-carrier embryo is at higher risk due to PRS.

2) By contrast, carriers of moderate penetrance (OR~2) risk variants for breast cancer (ATM, CHEK2, BARD1, RAD51C/D) demonstrated risk profiles overlapping non-carriers with PRS ~1.5SD higher (black dotted line).

1) Carriers of BRCA1 had risk distributions that were disjoint from non-carriers, regardless of polygenic risk scores (red dashed line in Figure above). Similar results were observed for BRCA2 and PALB2 carriers (blue dashed line).

Check out the paper for demonstrations of this effect in simulations and three datasets. We need to understand how PGS perform in clinical settings if they are ever to be useful as predictive markers, which is what we are trying to accomplish with the FEP-GEN consortium. 5/6

Are one or more of the polygenic embryo screening companies currently active/visible in South Korea?

See my pinned tweet for more details

Given that we are unlikely to scale much beyond this enormous dataset anytime soon, this appears to be quite a negative result…

Congratulations, Sunny—All the more impressive given the current environment. Your work is truly at the cutting edge!

This looks fascinating - I can’t wait to read it in detail.

I am wondering if your approach resembles something we tried on a much smaller scale a few years ago: pubmed.ncbi.nlm.nih.gov/32956347/

Need to double-check but I think it could be design / MAF cut-offs for each type of study

More details at the link:
t.co/ThiWw7QdkU

Concordant SNPs (low cognition / increased SZ risk) were associated with early neurodevelopmental (prenatal) forebrain abnormalities. Discordant SNPs (high cognition / increased SZ risk) associated with adult synaptic function and hindbrain (cerebellar) development. (2/3)

💯 always take the reader by the hand and walk them through slowly paragraph by paragraph

We identified 113 Bonferroni-corrected putatively causal associations (46 novel) involving 91 proteins, providing support for repurposing of anti-inflammatory agents for SCZ, amantadine for BD, retinoic acid for MDD, and duloxetine for CTP.
jamanetwork.com/journals/jam...