Sign in

Xilin Jiang

@xilinjiang.bsky.social
25 followers 25 following 9 posts

Wellcome Trust Fellow | Rhodes Scholar

PostsRepliesMedia
Xilin Jiang @xilinjiang.bsky.social · 03/09/2026
Led by postdoctoral fellow Pearl Liu; there is a lot more in www.medrxiv.org/content/10.6...
medrxiv.org
Predicting COVID-19 hospitalisation and common disease risk from comorbid diagnoses in 13 million individuals
Predicting disease risk from prior diagnoses is fundamental to clinical decision-making, particularly during health emergencies such as the COVID-19 pandemic, when individuals with long-term condition...
010 0
Xilin Jiang @xilinjiang.bsky.social · 03/09/2026
Lipid disorder is not predictive of Storke after conditioning on all prior diseases!
100 0
Xilin Jiang @xilinjiang.bsky.social · 03/09/2026
The model captures nearly all information of clinical comorbidity index and socio-economic status, just from prior diagnoses.
100 0
Xilin Jiang @xilinjiang.bsky.social · 03/09/2026
The model is also nearly perfectly transferable across self-reported ethnicities.
100 0
Xilin Jiang @xilinjiang.bsky.social · 03/09/2026
Surprisingly, we achieved a comparable prediction power (via the much larger sample sizes) than much more complicated models such as transformer
100 0
Xilin Jiang @xilinjiang.bsky.social · 03/09/2026
First, there is little overfitting as training R2 is getting very close to testing R2; an elastic-net model converges to a simple regression (“Bernstein–von Mises” in real life!)
100 0
Xilin Jiang @xilinjiang.bsky.social · 03/09/2026
We’ve trained a simple linear model using diagnosis records of all age 40-69 England population (13 million!) to estimate the prior diagnosis effect on future diseases. Here’s what we’ve found:
100 0
Xilin Jiang @xilinjiang.bsky.social · 03/09/2026
What happens when we have “sufficiently” large sample sizes? www.medrxiv.org/content/10.6...
medrxiv.org
Predicting COVID-19 hospitalisation and common disease risk from comorbid diagnoses in 13 million individuals
Predicting disease risk from prior diagnoses is fundamental to clinical decision-making, particularly during health emergencies such as the COVID-19 pandemic, when individuals with long-term condition...
111 0
Xilin Jiang @xilinjiang.bsky.social · 15/06/2026
Roughly half of common disease heritability is pleiotropic across diseases, with shared variance 1.51x enriched in the genetic component. Work led by Yujie Zhao (co-supervised by Alkes and me) is now out in @natgenet.nature.com www.nature.com/articles/s41...
051 0
Reposted by Xilin Jiang
Mike Inouye @mikeinouye.bsky.social · 09/06/2026
Huge congrats to @xilinjiang.bsky.social (Wellcome fellow in the group) on his first senior author paper, in @natgenet.nature.com no less! 👏👏👏👏 Pleiotropic shared heritability quantifies the shared genetic variance of common diseases www.nature.com/articles/s41...
nature.com
Pleiotropic shared heritability quantifies the shared genetic variance of common diseases - Nature Genetics
Pleiotropic shared heritability with bias correction (PHBC) estimates the genetic variance of a target disease that is shared with a set of auxiliary diseases. Applying PHBC to diseases in the UK Biob...
084 0
Reposted by Xilin Jiang
Arun Durvasula @arundurvasula.bsky.social · 03/09/2025
Excited to share our latest manuscript, "Exposure accumulation drives age-dependent disease architectures and polygenic risk scores," led by Xilin Jiang: www.medrxiv.org/content/10.1... I am attempting an explainer thread for the first time here: (I am usually too exhausted to post one)
medrxiv.org
Exposure accumulation drives age-dependent disease architectures and polygenic risk scores
Our understanding of the dependence of the genetic and environmental architecture of common diseases on age is incomplete. Here, we use longitudinal data to quantify age-dependent genetic and environm...
25425 0