Selin Jessa
@selinjessa.bsky.social
(she/her) Computational biologist and post-doc scientist in the Greenleaf and Kundaje labs at Stanford. Interested in understanding how cells know what to become (transcription factors, gene regulation, dev bio, open science) www.selinjessa.com
And then we stratified off-target base edits in non-coding loci based on their predicted consequences on the epigenome. In a case study, an intergenic off-target edit overlaps multiple motifs - our models predict that it specifically disrupts an AP-1 site. Much more in the paper, check out Tong's 🧵!
November 7, 2025 at 6:38 PM
And then we stratified off-target base edits in non-coding loci based on their predicted consequences on the epigenome. In a case study, an intergenic off-target edit overlaps multiple motifs - our models predict that it specifically disrupts an AP-1 site. Much more in the paper, check out Tong's 🧵!
We then used sequence-to-activity deep learning models, to predict effects of non-coding edits on TF binding and chromatin accessibility. We first show that a ChromBPNet model can predict the same GATA site disruption mechanism exploited by the FDA-approved Casgevy medicine, specifically in T cells:
November 7, 2025 at 6:38 PM
We then used sequence-to-activity deep learning models, to predict effects of non-coding edits on TF binding and chromatin accessibility. We first show that a ChromBPNet model can predict the same GATA site disruption mechanism exploited by the FDA-approved Casgevy medicine, specifically in T cells:
And then we stratified off-target base edits in non-coding loci based on their predicted consequences on the epigenome. We show a case study of an intergenic off-target edit overlapping multiple motifs. Our models predict that it disrupts an AP-1 site. So much more in the paper, check out Tong's 🧵!
November 7, 2025 at 4:25 PM
And then we stratified off-target base edits in non-coding loci based on their predicted consequences on the epigenome. We show a case study of an intergenic off-target edit overlapping multiple motifs. Our models predict that it disrupts an AP-1 site. So much more in the paper, check out Tong's 🧵!
We then used sequence-to-activity deep learning models, to predict effects of non-coding edits on TF binding and chromatin accessibility. We first show that a ChromBPNet model can predict the same GATA site disruption mechanism exploited by the FDA-approved Casgevy medicine, specifically in T cells:
November 7, 2025 at 4:25 PM
We then used sequence-to-activity deep learning models, to predict effects of non-coding edits on TF binding and chromatin accessibility. We first show that a ChromBPNet model can predict the same GATA site disruption mechanism exploited by the FDA-approved Casgevy medicine, specifically in T cells:
Last, we found that putative causal noncoding variants for various diseases were enriched in cREs in a cell type-specific manner, and we used our models to predict how single nucleotide changes disrupt/create motifs and thus alter accessibility, providing mechanistic interpretation of variant effect
May 3, 2025 at 6:27 PM
Last, we found that putative causal noncoding variants for various diseases were enriched in cREs in a cell type-specific manner, and we used our models to predict how single nucleotide changes disrupt/create motifs and thus alter accessibility, providing mechanistic interpretation of variant effect
Deep learning models can learn which sequences are predictive of accessibility, but also which *negatively* influence accessibility. We discovered a handful of negative motifs (in peaks!) which were extremely abundant in every cell type, enriched near nucleosome dyads, & concentrated at peak flanks
May 3, 2025 at 6:27 PM
Deep learning models can learn which sequences are predictive of accessibility, but also which *negatively* influence accessibility. We discovered a handful of negative motifs (in peaks!) which were extremely abundant in every cell type, enriched near nucleosome dyads, & concentrated at peak flanks
We also found examples of "soft" syntax, where motifs synergize across longer distances (<150 bp), potentially reflecting active or passive competition of TFs with nucleosomes mediating cooperativity:
May 3, 2025 at 6:27 PM
We also found examples of "soft" syntax, where motifs synergize across longer distances (<150 bp), potentially reflecting active or passive competition of TFs with nucleosomes mediating cooperativity:
Our models were able to predict synergistic effects at exactly the motif syntax described for the Coordinator motif by @seungsookim.bsky.social/Wysocka lab, where X-ray crystallography showed DNA facilitates weak contacts between TWIST1 & ALX4, and the TF complex directs mesenchymal gene programs
May 3, 2025 at 6:27 PM
Our models were able to predict synergistic effects at exactly the motif syntax described for the Coordinator motif by @seungsookim.bsky.social/Wysocka lab, where X-ray crystallography showed DNA facilitates weak contacts between TWIST1 & ALX4, and the TF complex directs mesenchymal gene programs
We identified 100s of composite motifs, so we used our models to run in silico experiments to systematically define effects of motif syntax (spacing/orientation) on synergy. We found dozens of cases of "hard" syntax, where synergy relies on strict motif position, likely due to direct interactions:
May 3, 2025 at 6:27 PM
We identified 100s of composite motifs, so we used our models to run in silico experiments to systematically define effects of motif syntax (spacing/orientation) on synergy. We found dozens of cases of "hard" syntax, where synergy relies on strict motif position, likely due to direct interactions:
We used this map of motif instances in every cell type to identify ubiquitous and cell type specific motifs, and found that a small set of ubiquitous, CG-rich motifs tended to occur in promoters, while cell type specific motifs were predominant at distal and intronic regions:
May 3, 2025 at 6:27 PM
We used this map of motif instances in every cell type to identify ubiquitous and cell type specific motifs, and found that a small set of ubiquitous, CG-rich motifs tended to occur in promoters, while cell type specific motifs were predominant at distal and intronic regions:
Clustering these motifs, we assembled a lexicon of 508 unique motifs which influence accessibility during development, and mapped these back to the peak regions to automatically annotate predictive motif instances in open chromatin in every cell type, representing putative TF binding sites
May 3, 2025 at 6:27 PM
Clustering these motifs, we assembled a lexicon of 508 unique motifs which influence accessibility during development, and mapped these back to the peak regions to automatically annotate predictive motif instances in open chromatin in every cell type, representing putative TF binding sites
We used model interpretation techniques (DeepLIFT/SHAP & TF-MoDISco) to score the contribution of every nucleotide to accessibility, and discover recurrent patterns of sequences predictive of local chromatin accessibility - and these patterns turned out to primarily resemble TF binding motifs!
May 3, 2025 at 6:27 PM
We used model interpretation techniques (DeepLIFT/SHAP & TF-MoDISco) to score the contribution of every nucleotide to accessibility, and discover recurrent patterns of sequences predictive of local chromatin accessibility - and these patterns turned out to primarily resemble TF binding motifs!
What are the DNA sequences that drive accessibility in each cell type? In every cell type, we trained ChromBPNet models - deep learning models tasked with predicting chromatin accessibility in 1 kbp regions at basepair resolution from 2 kbp local sequence alone. These models work remarkably well:
May 3, 2025 at 6:27 PM
What are the DNA sequences that drive accessibility in each cell type? In every cell type, we trained ChromBPNet models - deep learning models tasked with predicting chromatin accessibility in 1 kbp regions at basepair resolution from 2 kbp local sequence alone. These models work remarkably well:
Do these cREs drive activity in vivo? We inspected the VISTA enhancers (validated in reporter mice), and our data suggested previously unappreciated activity of some enhancers in the liver! Our accessibility data resolved specificity of one enhancer to erythroblasts, which we confirmed w/ histology:
May 3, 2025 at 6:27 PM
Do these cREs drive activity in vivo? We inspected the VISTA enhancers (validated in reporter mice), and our data suggested previously unappreciated activity of some enhancers in the liver! Our accessibility data resolved specificity of one enhancer to erythroblasts, which we confirmed w/ histology:
We generated a single-cell multi-ome, multi-organ atlas of human development using SHARE-seq, profiling gene expression & accessibility in 818k cells from 12 organs, 10-23 PCW: the Human Development Multiomic Atlas (HDMA). We annotated 203 cell types & defined >1M candidate cis-regulatory elements:
May 3, 2025 at 6:27 PM
We generated a single-cell multi-ome, multi-organ atlas of human development using SHARE-seq, profiling gene expression & accessibility in 818k cells from 12 organs, 10-23 PCW: the Human Development Multiomic Atlas (HDMA). We annotated 203 cell types & defined >1M candidate cis-regulatory elements:
Last, we found that putative causal noncoding variants for various diseases were enriched in cREs in a cell type-specific manner, and we used our models to predict how single nucleotide changes disrupt/create motifs and thus alter accessibility, providing mechanistic interpretation of variant effect
May 3, 2025 at 6:02 PM
Last, we found that putative causal noncoding variants for various diseases were enriched in cREs in a cell type-specific manner, and we used our models to predict how single nucleotide changes disrupt/create motifs and thus alter accessibility, providing mechanistic interpretation of variant effect
Deep learning models can learn which sequences are predictive of accessibility, but also which *negatively* influence accessibility. We discovered a handful of negative motifs (in peaks!) which were extremely abundant in every cell type, enriched near nucleosome dyads, & concentrated at peak flanks
May 3, 2025 at 6:02 PM
Deep learning models can learn which sequences are predictive of accessibility, but also which *negatively* influence accessibility. We discovered a handful of negative motifs (in peaks!) which were extremely abundant in every cell type, enriched near nucleosome dyads, & concentrated at peak flanks
We also found examples of "soft" syntax, where motifs synergize across longer distances (<150 bp) with decaying effects as they move further apart, potentially reflecting active or passive competition of TFs with nucleosomes mediating cooperativity:
May 3, 2025 at 6:02 PM
We also found examples of "soft" syntax, where motifs synergize across longer distances (<150 bp) with decaying effects as they move further apart, potentially reflecting active or passive competition of TFs with nucleosomes mediating cooperativity:
For example, our models were able to predict synergistic effects at exactly the motif syntax described for the Coordinator motif by @seungsookim & Wysocka lab, where X-ray crystallography showed DNA facilitates weak contacts between TWIST1 & ALX4, and the TF complex directs mesenchymal gene programs
May 3, 2025 at 6:02 PM
For example, our models were able to predict synergistic effects at exactly the motif syntax described for the Coordinator motif by @seungsookim & Wysocka lab, where X-ray crystallography showed DNA facilitates weak contacts between TWIST1 & ALX4, and the TF complex directs mesenchymal gene programs
We identified 100s of composite motifs, so we used our models to run in silico experiments to systematically define effects of motif syntax (spacing/orientation) on synergy. We found 48 cases of "hard" syntax, where synergy relies on strict motif position, likely due to direct protein interactions:
May 3, 2025 at 6:02 PM
We identified 100s of composite motifs, so we used our models to run in silico experiments to systematically define effects of motif syntax (spacing/orientation) on synergy. We found 48 cases of "hard" syntax, where synergy relies on strict motif position, likely due to direct protein interactions:
We used this map of motif instances in every cell type to identify ubiquitous and cell type specific motifs, and found that a small set of ubiquitous, CG-rich motifs tended to occur in promoters, while cell type-specific motifs were predominant at distal and intronic regions:
May 3, 2025 at 6:02 PM
We used this map of motif instances in every cell type to identify ubiquitous and cell type specific motifs, and found that a small set of ubiquitous, CG-rich motifs tended to occur in promoters, while cell type-specific motifs were predominant at distal and intronic regions:
Clustering these motifs, we assembled a lexicon of 508 unique motifs which influence accessibility during development, and mapped these back to the peak regions to automatically annotate predictive motif instances in open chromatin in every cell type, representing putative TF binding sites:
May 3, 2025 at 6:02 PM
Clustering these motifs, we assembled a lexicon of 508 unique motifs which influence accessibility during development, and mapped these back to the peak regions to automatically annotate predictive motif instances in open chromatin in every cell type, representing putative TF binding sites:
We used model interpretation techniques (DeepLIFT/SHAP & TF-MoDISco) to score the contribution of every nucleotide to accessibility, and discover recurrent patterns of sequences predictive of local chromatin accessibility - and these patterns turned out to primarily resemble TF binding motifs!
May 3, 2025 at 6:02 PM
We used model interpretation techniques (DeepLIFT/SHAP & TF-MoDISco) to score the contribution of every nucleotide to accessibility, and discover recurrent patterns of sequences predictive of local chromatin accessibility - and these patterns turned out to primarily resemble TF binding motifs!
What are the DNA sequences that drive accessibility in each cell type? In every cell type, we first trained ChromBPNet models - deep learning models tasked with predicting chromatin accessibility in 1 kbp regions at basepair resolution from 2 kbp local sequence alone, resulting in a suite of models:
May 3, 2025 at 6:02 PM
What are the DNA sequences that drive accessibility in each cell type? In every cell type, we first trained ChromBPNet models - deep learning models tasked with predicting chromatin accessibility in 1 kbp regions at basepair resolution from 2 kbp local sequence alone, resulting in a suite of models:
Do these cREs drive activity in vivo? We inspected the VISTA enhancers (validated in reporter mice), and our data suggested previously unappreciated activity of some enhancers in the liver! Our accessibility data resolved specificity of one enhancer to erythroblasts, which we confirmed w/ histology:
May 3, 2025 at 6:02 PM
Do these cREs drive activity in vivo? We inspected the VISTA enhancers (validated in reporter mice), and our data suggested previously unappreciated activity of some enhancers in the liver! Our accessibility data resolved specificity of one enhancer to erythroblasts, which we confirmed w/ histology: