DeepSTARR predicts enhancer activity from DNA sequence and enables the<i>de novo</i>design of enhancers

Almeida, Bernardo Pinto de; Reiter, Franziska; Pagani, Michaela; Stark, Alexander

doi:10.1101/2021.10.05.463203

Cited by 12 publications

(13 citation statements)

References 117 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Strikingly, we find that flanking nucleotide combinations can drive predictions by a factor of 3 relative to the AP-1’s core binding site with random flanks (Fig. 6b), similar to what was observed previously 45,46 . A position-weight-matrix-based approach 47 , which considers each position independently, would score many AP-1 binding sites the same, despite their wide spread in functional activity.…”

Section: Resultssupporting

confidence: 88%

Evaluating deep learning for predicting epigenomic profiles

Toneyan

Tang

Koo

2022

Preprint

View full text Add to dashboard Cite

Deep learning has been successful at predicting epigenomic profiles from DNA sequences. Most approaches frame this task as a binary classification relying on peak callers to define functional activity. Recently, quantitative models have emerged to directly predict the experimental coverage values as a regression. As new models continue to emerge with different architectures and training configurations, a major bottleneck is forming due to the lack of ability to fairly assess the novelty of proposed models and their utility for downstream biological discovery. Here we introduce a unified evaluation framework and use it to compare various binary and quantitative models trained to predict chromatin accessibility data. We highlight various modeling choices that affect generalization performance, including a downstream application of predicting variant effects. In addition, we introduce a robustness metric that can be used to enhance model selection and improve variant effect predictions. Our empirical study largely supports that quantitative modeling of epigenomic profiles leads to better generalizability and interpretability.

show abstract

Section: Resultssupporting

confidence: 88%

Evaluating deep learning for predicting epigenomic profiles

Toneyan

Tang

Koo

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Fig. S2: Cooperativity (residual fold change; y-axis) plotted as a function of distance (x-axis) between the motifs of the housekeeping TFs Dref (top row), Ohler1 (middle row), and Ohler6 (bottom row) for DeepSTARR 72 (left column) and ExplaiNN (right column). The 5-mer GGGCT is provided as a negative control (blue).…”

Section: Figuresmentioning

confidence: 99%

ExplaiNN: interpretable and transparent neural networks for genomics

Novakovsky

Fornés

Saraswat

et al. 2022

Preprint

View full text Add to dashboard Cite

Sequence-based deep learning models, particularly convolutional neural networks (CNNs), have shown superior performance on a wide range of genomic tasks. A key limitation of these models is the lack of interpretability, slowing their broad adoption by the genomics community. Current approaches to model interpretation do not readily reveal how a model makes predictions, can be computationally intensive, and depend on the implemented architecture. Here, we introduce ExplaiNN, an adaptation of neural additive models1 for genomic tasks wherein predictions are computed as a linear combination of multiple independent CNNs, each consisting of a single convolutional filter and fully connected layers. This approach brings together the expressivity of CNNs with the interpretability of linear models, providing global (cell state level) as well as local (individual sequence level) insights of the biological processes studied. We use ExplaiNN to predict transcription factor (TF) binding and chromatin accessibility states, demonstrating performance levels comparable to state-of-the-art methods, while providing a transparent view of the model’s predictions in a straightforward manner. Applied to de novo motif discovery, ExplaiNN detects equivalent motifs to those obtained from specialized algorithms across a range of datasets. Finally, we present ExplaiNN as a plug and play platform in which pre-trained TF binding models and annotated position weight matrices from reference databases can be combined in a simple framework. We expect that ExplaiNN will accelerate the adoption of deep learning by biological domain experts in their daily genomic sequence analyses.

show abstract

“…Additional mechanistic insight has been provided by thermodynamic modelling of enhancers 34,35 , in vivo imaging of enhancer activity 36 , the analysis of genetic variation through eQTL and caQTL analysis 6,37 , and high-throughput in vitro binding assays 38,39 . Recently, the enhancer biology field embraced the use of convolutional neural networks (CNN) and network-explainability techniques that again provided a significant leap forward in terms of prediction accuracy and syntax formulation 10,[40][41][42][43][44] . An orthogonal strategy to decode enhancer logic is to engineer synthetic enhancers from scratch.…”

Section: Introductionmentioning

confidence: 99%

“…This approach has the advantage that the designer knows exactly which features are implanted, so that the minimal requirements for enhancer function can be revealed. Recent work showed the promise of CNN-driven enhancer design by successfully designing yeast promoters 45 , and by using a CNN to select high-scoring enhancers for S2 cells, from a large pool of random sequences 42 . Here we tackle the next challenge in enhancer design, namely to design enhancers that are cell type specific.…”

Section: Introductionmentioning

confidence: 99%

Cell type directed design of synthetic enhancers

Taskiran

Spanier

Christiaens

et al. 2022

Preprint

View full text Add to dashboard Cite

Transcriptional enhancers act as docking stations for combinations of transcription factors and thereby regulate spatiotemporal activation of their target genes. A single enhancer, of a few hundred base pairs in length, can autonomously and independently of its location and orientation drive cell-type specific expression of a gene or transgene. It has been a long-standing goal in the field to decode the regulatory logic of an enhancer and to understand the details of how spatiotemporal gene expression is encoded in an enhancer sequence. Recently, deep learning models have yielded unprecedented insight into the enhancer code, and well-trained models are reaching a level of understanding that may be close to complete. As a consequence, we hypothesized that deep learning models can be used to guide the directed design of synthetic, cell type specific enhancers, and that this process would allow for a detailed tracing of all enhancer features at nucleotide-level resolution. Here we implemented and compared three different design strategies, each built on a deep learning model: (1) directed sequence evolution; (2) directed iterative motif implanting; and (3) generative design. We evaluated the function of fully synthetic enhancers to specifically target Kenyon cells in the fruit fly brain using transgenic animals. We then exploited this concept further by creating "dual-code" enhancers that target two cell types, and minimal enhancers smaller than 50 base pairs that are fully functional. By examining the trajectories followed during state space searches towards functional enhancers, we could accurately define the enhancer code as the optimal strength, combination, and relative distance of TF activator motifs, and the absence of TF repressor motifs. Finally, we applied the same three strategies to successfully design human enhancers. In conclusion, enhancer design guided by deep learning leads to better understanding of how enhancers work and shows that their code can be exploited to manipulate cell states.

show abstract

DeepSTARR predicts enhancer activity from DNA sequence and enables thede novodesign of enhancers

Cited by 12 publications

References 117 publications

Evaluating deep learning for predicting epigenomic profiles

Evaluating deep learning for predicting epigenomic profiles

ExplaiNN: interpretable and transparent neural networks for genomics

Cell type directed design of synthetic enhancers

Contact Info

Product

Resources

About