Brilliant news, very well done!

Most importantly, thank you to past/present team members for their huge efforts- Elena Arciero, @emiliewigdor.bsky.social, Mari Niemi, @chundru.bsky.social , Patrick Campbell, Teng Heng, Daniel Malawsky, @qinqinhuang.bsky.social, @yiorkala.bsky.social, @olivia-wootton.bsky.social, Klaudia Walter.

Many thanks to @gensocuk.bsky.social for this honour & recognition of my team's work, & to @sangerinstitute.bsky.social for the huge opportunity it has given me. Very grateful to @mehurles.bsky.social , @jeffbarrett.eu, @carlanderson.bsky.social , Peter Donnelly and Gil McVean for their support.

Thanks, @wendydjones.bsky.social!

Huge thanks to my supervisors @r-rahbari.bsky.social & @hilarycmartin.bsky.social, our collaborators, and most importantly, the @genomicsengland.bsky.social participants for making this research possible! ✨✨

It was indeed a great conference, I really enjoyed it and would highly recommend to others! Excellent location too.

A key methodological advantage is the use of Mendelian imputation to recover missing parental genotypes. This helps mitigate bias associated with non-random ascertainment of genotyped trios in these cohorts, over and above boosting power for association testing (See Extended Data Figure 1 and 2).

I’m most excited about our rare variant findings: a higher burden of deleterious protein-truncating variants is associated with increased mental health symptoms. Trio models indicated direct genetic effects on externalising in MCS and on internalising symptoms in ALSPAC.

Thanks to all co-authors particularly co-first authors David van den Berg (Abdel's group), Wei Huang (Matt's group) and Daniel Malawsky (my group) - none of whom are yet on bluesky!

Finally, importantly, we replicate (in aggregate) the new genes showing rare variant associations with FI in independent cohorts with direct measures of cognitive ability (ALSPAC, MCS) & show that the ones not previously linked with NDDs are enriched ~2.5-fold for de novo PTVs in 31k NDD patients.

One cool finding: rare protein-truncating variants (PTVs) in CADPS2 are signif associated with *increased* EA (most genes show negative assocs) but no signal for FI. This is consistent with common variant findings at the same locus (Supp Note 7). Maybe a non-cognitive mechanism? Replication needed!

We show that using measured+imputed FI gives more power in rare variant tests than using educational attainment (EA) on the same sample (26 versus 9 signif genes). (cont)

Other new genes with signif rare associations include some excellent candidates (e.g. ROBO2, CHD9) and some wild-cards (DPP8, IPO9) - maybe some novel biology there? (cont)

I'm most excited about the rare variant findings: 26 genes with associations to measured+imputed fluid intelligence (FI), of which only 5 would be found using measured FI alone. 14/26 are well-known neurodevelopmental disorder (NDD) genes (4-fold enrichment, p=8e-8). (cont)

For a lay-language summary and FAQ containing additional context, see the Supp www.medrxiv.org/content/10.1.... Thanks to @ewanbirney.bsky.social for his help on this.