Alexander Stark

The paper designs tissue-specific enhancers for mouse embryos using deep learning models trained on chromatin accessibility data and validated enhancers.

BARe-seq measures transcriptional burst parameters in Drosophila cells using allele-resolved bulk sequencing, revealing distinct kinetic properties of promoters and enhancers.

The paper systematically identifies effector functions of pathogen ORFs across multiple pathways, revealing new roles for known effectors and uncharacterized proteins.

The paper identifies Hcfc1 as a CGI-specific co-activator that bridges transcription factors to TFIID, revealing a promoter-class-specific activation mechanism for CGI promoters.

NTv3 is a foundation model that predicts genome function, annotates genes, and designs DNA sequences across species, outperforming existing models in multiple tasks.

The paper introduces Ledidi, a computational method for designing cis-regulatory elements by optimizing continuous sequences guided by deep learning models, validated experimentally for controlling transcriptional outcomes.

The paper shows that enhancer cooperativity can compensate for reduced activity over large genomic distances by introducing weak enhancers between strong enhancers and promoters.

The paper investigates enhancer-enhancer cooperativity in Drosophila, finding that housekeeping enhancers are additive while developmental enhancers are synergistic.

ORFtag enables proteome-scale tagging and functional screening in mammalian cells, identifying known and novel regulators of transcription and post-transcriptional processes.

This study identifies TF motifs that determine enhancer sensitivity or resistance to specific co-repressors in Drosophila, revealing regulatory rules for repression mechanisms.

The paper shows that different promoter classes use distinct PIC compositions and mechanisms, leading to focused or dispersed transcription initiation patterns in Drosophila.

The paper investigates enhancer sequence flexibility and context-specific motif function in Drosophila, finding constrained flexibility and context-dependent motif activity.

The paper identifies and characterizes repressive domains in Drosophila transcription factors using a new high-throughput method, revealing conserved motifs and co-repressor interactions.

This paper introduces DeepSTARR, a deep learning model that predicts enhancer activity from DNA sequence and enables the de novo design of enhancers in Drosophila and human.

The paper identifies and corrects systematic errors in enhancer activity assays, improving the reliability of genome-wide enhancer characterization in human cells.

The paper presents a genome-wide Ubx binding map in Drosophila using ChIP-seq, showing clustered binding sites at developmental regulators and a potential link to Polycomb-mediated regulation.