Leor S. Weinberger

The paper critiques the validity of using observational data to infer the antiviral efficacy of therapeutic interfering particles (TIPs) and highlights inconsistencies in reported viral dynamics.

CONCORD is a self-supervised framework that integrates single-cell datasets across batches, technologies, and species, producing high-resolution cell atlases with improved biological resolution and reduced noise.

This paper compares scRNA-seq and smFISH for quantifying transcriptional noise, finding that scRNA-seq algorithms generally underestimate noise compared to smFISH.

The paper compares RNA-seq and single-molecule RNA FISH to show that IdU globally amplifies transcriptional noise across most genes.

The paper identifies DNA base-excision repair as a mechanism that amplifies transcriptional noise, enhancing cellular plasticity and reprogramming efficiency.

RanDeL-seq maps viral cis- and trans-acting elements by generating and screening high-throughput deletion mutants, revealing critical regions in HIV-1 and ZIKV.

The paper identifies stochastic variability in viral tegument protein abundance as a mechanism for probabilistic bet hedging between viral replication and latency in herpesvirus.

This study shows that Brm deletion causes identity switches in cardiac mesoderm differentiation by altering chromatin accessibility and gene expression, revealing BRM's role in safeguarding developmental canalization.

This paper discovers and engineers an HIV-1 Therapeutic Interfering Particle (TIP) that suppresses viral replication and protects CD4+ T cells in humanized mice.

The paper shows that positive feedback in HCMV IE1 protein sustains lytic gene expression despite transient tegument transactivator signals, enhancing viral fitness.

The paper investigates how HIV's Tat protein generates bimodal expression through non-latching positive feedback, decoupling noise from mean expression to maintain stochastic bimodality.

The paper re-examines HIV LTR mRNA and protein expression data, finding that noise scales inversely with mean expression, supporting the inverse correlation between noise and mean.