Thread by @Jacob_Vorstman

1/X rdcu.be/b992F
We are very excited about our study appearing today in @NatureMedicine. It is a genotype-phenotype study in the largest sample of individuals with #22q11.2 deletion syndrome reported to date.

Let us give you a quick 12-step tour of our findings.

2/X #22q11DS is a #RareDisease. In addition to a range of physical symptoms, patients are liable to neuropsychiatric conditions, including #schizophrenia (SZ) and #intellectualdisability (ID). However, manifestations are highly variable. Image: positiveexposure.org

3/X With the International 22q11 Brain and Behavior Consortium (IBBC), we examined the role of common genetic variation, summarized in a polygenic score (PS) in the context of the 22q11.2 deletion. @NIMHgov @WiringTheBrain @EricTopol @22qFoundation @UCLAHealth

4/X IQ is normally distributed in 22q11DS, with a ~30pts left shift attributed to the impact of the deletion. We hypothesized that variance in IQ in 22q11DS is likely also under the same influence of common variants (just like in the Gen Pop). @cardiffuni @UMCUtrecht

5/X In our sample of 965 individuals with 22q11DS, we showed that this was indeed the case, both for IQ and for Schizophrenia. Both phenotypes were associated with polygenic scores, echoing the same observations in the general population. @AFiksinski @robbie_stats @Winkytheelf

6/X Why is this relevant? These findings are consistent with a model in which common risk variation helps explain some of the variable penetrance observed in virtually all rare pathogenic variants of which 22q11DS is an example. @UniofOxford @sebatlab @PennMedNews @PennMedicine

7/X A previous IBBC study demonstrated that a subgroup of 22q11DS has a loss of cognitive skills in excess of the moderate avg decline observed on the group level. This subset was found to be at further increased risk for schizophrenia spectrum disorders. @JAMAPsych @KU_Leuven

8/X We studied the genetic underpinnings of verbal IQ (VIQ) decline and subthreshold psychosis, both early schizophrenia-related manifestations. We found PS_SZ, but not PS_IQ, was associated with VIQ decline, and a nominally signif. assoc. between PS_SZ and subthr. psychosis.

9/X The implication of these observations is a shared genetic basis between and the early schizophrenia-related phenotypes of cognitive decline and (tentatively) positive psychotic symptoms. @gmrepetto @CALDirector @jongepsychiater @RCSI_Research

10/X Lastly, we examined if info from common risk variants in "the rest of the genome", summarized as polygenic score, can be used for individual risk prediction. Different Odds Ratios can be calculated as a function of the percentile used as a cut-off in the PS distribution.

11/X In the gen pop, even the extremities of the distribution do not amount to informative prediction metrics (Positive Prediction Values). In contrast, in 22q11DS the same ORs act upon high pre-test probabilities, which generates PPVs approaching the realm of clinical utility.

12/X
In sum, we show:
I. Role of common variation in explaining variable penetrance of 22q11DS
II. Genetic underpinnings of early SZ-related phenotypes
III. Potential of polygenic scores for individual risk prediction when applied in a priori high risk samples (e.g. 22q11DS)

Thank you for your interest. You can read more here: rdcu.be/b992F and doi.org/10.1038/s41591…
Or here: sickkids.ca/AboutSickKids/…