BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Knowledge Exchange: Causal Network Recovery in Perturb-seq Experim
 ents Using Proxy and Instrumental Variables - Kwangmoon Park (University o
 f Pennsylvania)
DTSTART:20260311T100000Z
DTEND:20260311T110000Z
UID:TALK245617@talks.cam.ac.uk
DESCRIPTION: Emerging single-cell technologies that integrate CRISPR-based
  genetic perturbations with single-cell RNA sequencing\, such as Perturb-s
 eq\, have substantially advanced our understanding of gene regulation and 
 causal influence of genes. While Perturb-seq data provide valuable causal 
 insights into gene-gene interactions\, statistical concerns remain regardi
 ng unobserved confounders that may bias inference. These latent factors ma
 y arise not only from intrinsic molecular features of regulatory elements 
 encoded in Perturb-seq experiments\, but also from unobserved genes arisin
 g from cost-constrained experimental designs. Although methods for analyzi
 ng large-scale Perturb-seq data are rapidly maturing\, approaches that exp
 licitly account for such unobserved confounders in learning the causal gen
 e networks are still lacking. Here\, we propose a novel method to recover 
 causal gene networks from Perturb-seq experiments with robustness to arbit
 rarily omitted confounders. Our framework leverages proxy and instrumental
  variable strategies to exploit the rich information embedded in perturbat
 ions\, enabling unbiased estimation of the underlying directed acyclic gra
 ph (DAG) of gene expressions. Simulation studies and analyses of CRISPR in
 terference experiments of K562 cells demonstrate that our method outperfor
 ms baseline approaches that ignore unmeasured confounding\, yielding more 
 accurate and biologically relevant recovery of the true gene causal DAGs.
LOCATION:Seminar Room 2\, Newton Institute
END:VEVENT
END:VCALENDAR