Deep learning-based enhancement of epigenomics data with AtacWorks

A, Lal; Chiang, Zachary; Yakovenko, Nikolai; Duarte, Fabiana M.; Israeli, Johnny; Buenrostro, Jason D.

doi:10.1038/s41467-021-21765-5

Cited by 41 publications

(21 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ATAC-seq data sets from cryopreserved tissue were slightly lower quality compared to data sets obtained from fresh tissue, a known issue when processing frozen samples and tissues specifically [31]. Processing with the AtacWorks algorithm [36] resulted in improved signal quality, similar to what we observed with processed data from fresh tissue ( Figure 5B ).…”

Section: Resultssupporting

confidence: 74%

“…We observed enrichment of open chromatin at known oocyte-specific and granulosa cell-specific markers, GDF-9 and FOXL-2, respectively, that appeared consistent between donors ( Figure 4C ), indicative of isolation of the cell types of interest. To improve the signal over background in our data sets, we used a deep learning toolkit called AtacWorks [36] to denoise our sequencing data ( Figure 4D ).…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Application of Omni-ATAC to Profile Chromatin Accessibility Before and After Ovarian Tissue Cryopreservation

Shannon

Sundaresan

Bükülmez

et al. 2021

Preprint

View full text Add to dashboard Cite

Ovarian tissue cryopreservation and subsequent autologous transplantation has allowed resumption of endocrine function as well as fertility in certain populations. However, graft function is short-lived due to ischemia and aberrant follicular activation post-transplantation. While many studies have focused on gene expression, we wanted to determine whether cryopreservation itself had a deleterious effect on regulatory elements that might influence transcriptional integrity and graft performance. In this study, we used Omni-ATAC to assess genome-wide chromatin accessibility in primary human follicles before and after cryopreservation. Omni-ATAC from fresh ovarian follicles identified active regulatory elements expected to be functional in oocytes and granulosa cells, and gene ontology was consistent with RNA translation/processing and DNA repair. While promoter accessibility was largely maintained in cryopreserved ovarian follicles, we observed a widespread increase in the number of accessible enhancers. Transcription factor motif analysis and gene ontology suggested that this dysregulation was focused around the epithelial-mesenchymal transition. Indeed, transcription factor binding was noted in major pathways involved in this transition: TGF-β and Wnt signaling. Overall, our work provides the first genomic analysis of active regulatory elements in matched fresh and cryopreserved ovarian follicles as they undergo the process of ovarian tissue cryopreservation. Our data suggest that the process of cryopreservation activates an epithelial-mesenchymal transition state, which may lead to graft burn-out post-transplantation. Optimizing this technique in relation to this transition may therefore be an important step towards improving graft longevity and patient outcomes in fertility preservation.Summary sentenceCryopreservation of ovarian cortical tissue results in activation of differentiation and EMT pathways in follicles, which may explain graft burnout after autotransplantation.

show abstract

Section: Resultssupporting

confidence: 74%

Section: Resultsmentioning

confidence: 99%

Application of Omni-ATAC to Profile Chromatin Accessibility Before and After Ovarian Tissue Cryopreservation

Shannon

Sundaresan

Bükülmez

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Methods such as TF-MoDISco could be applied to scBasset ISM scores for de novo motif discovery (Shrikumar et al, 2018; Avsec et al, 2021). All approaches to scATAC analysis depend on accurate peak calls, and predictive modeling frameworks have been proposed to help identify highly specific regulatory elements (Lal et al, 2021). We expect a neural network model would further improve scATAC peak calling by taking into account sequence information (and accounting for Tn5 transposition bias).…”

Section: Discussionmentioning

confidence: 99%

scBasset: Sequence-based modeling of single cell ATAC-seq using convolutional neural networks

Yuan

Kelley

2021

Preprint

View full text Add to dashboard Cite

Single cell ATAC-seq (scATAC) shows great promise for studying cellular heterogeneity in epigenetic landscapes, but there remain significant challenges in the analysis of scATAC data due to the inherent high dimensionality and sparsity. Here we introduce scBasset, a sequence-based convolutional neural network method to model scATAC data. We show that by leveraging the DNA sequence information underlying accessibility peaks and the expressiveness of a neural network model, scBasset achieves state-of-the-art performance across a variety of tasks on scATAC and single cell multiome datasets, including cell type identification, scATAC profile denoising, data integration across assays, and transcription factor activity inference.

show abstract

“…Our work demonstrates the feasibility of training such complex models (thousands of free parameters) on limited data sets (hundreds rather than thousands of samples), and we have tested that it can handle other data sets of much larger size (tens of thousands of samples, data not shown). It stands in contrast to other available implementations of neural networks for regulatory genomics, which are targeted to modeling epigenomic ( 39 , 63 , 64 ) and genome-wide TF-DNA binding ( 36 , 38 ) data, or do not explicitly model the dependence of sequence function on cellular descriptors such as TF levels ( 65 ). This feature allows CoNSEPT to make predictions for varying cellular conditions.…”

Section: Discussionmentioning

confidence: 99%

Deciphering enhancer sequence using thermodynamics-based models and convolutional neural networks

Dibaeinia

Sinha

2021

Nucleic Acids Research

View full text Add to dashboard Cite

Deciphering the sequence-function relationship encoded in enhancers holds the key to interpreting non-coding variants and understanding mechanisms of transcriptomic variation. Several quantitative models exist for predicting enhancer function and underlying mechanisms; however, there has been no systematic comparison of these models characterizing their relative strengths and shortcomings. Here, we interrogated a rich data set of neuroectodermal enhancers in Drosophila, representing cis- and trans- sources of expression variation, with a suite of biophysical and machine learning models. We performed rigorous comparisons of thermodynamics-based models implementing different mechanisms of activation, repression and cooperativity. Moreover, we developed a convolutional neural network (CNN) model, called CoNSEPT, that learns enhancer ‘grammar’ in an unbiased manner. CoNSEPT is the first general-purpose CNN tool for predicting enhancer function in varying conditions, such as different cell types and experimental conditions, and we show that such complex models can suggest interpretable mechanisms. We found model-based evidence for mechanisms previously established for the studied system, including cooperative activation and short-range repression. The data also favored one hypothesized activation mechanism over another and suggested an intriguing role for a direct, distance-independent repression mechanism. Our modeling shows that while fundamentally different models can yield similar fits to data, they vary in their utility for mechanistic inference. CoNSEPT is freely available at: https://github.com/PayamDiba/CoNSEPT.

show abstract

Deep learning-based enhancement of epigenomics data with AtacWorks

Cited by 41 publications

References 43 publications

Application of Omni-ATAC to Profile Chromatin Accessibility Before and After Ovarian Tissue Cryopreservation

Application of Omni-ATAC to Profile Chromatin Accessibility Before and After Ovarian Tissue Cryopreservation

scBasset: Sequence-based modeling of single cell ATAC-seq using convolutional neural networks

Deciphering enhancer sequence using thermodynamics-based models and convolutional neural networks

Contact Info

Product

Resources

About