DeepBend: An interpretable model of DNA bendability

Khan, Samin Rahman; Sakib, Sadman; Rahman, Md. Saidur; Samee, Md. Abul Hassan

doi:10.1016/j.isci.2023.105945

Cited by 6 publications

(6 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both shuffling-type experiments indicate that dinucleotide composition plays an important role in the intrinsic cyclizability of a given DNA sequence. This agrees with previous studies that linked dinucleotides as key determinants of certain DNA mechanical properties (45,46) and the study by Khan et al (30), which made similar observations. When comparing the cyclizability signal to the distribution of all 16 dinucleotides at each position normalized to their random expectation, we found that their relative frequencies change markedly around the TSS and that TA frequency, in particular, correlated strongly with the cyclizability pattern found in Arabidopsis (Figure 5).…”

Section: Autocorrelation Of Cyclizability Shows No Large-scale Patternssupporting

confidence: 93%

“…Previous studies concluded that base composition is not a good indicator of DNA rigidity, except for mono-base or two-base repeats (44). This was confirmed by the two existing computational models (29,30) as well as by our model CycPred. Randomly shuffling random sequences (each base assumed with equal frequency (25%), shuffling at the level of individual bases) leads to a large variance of predictions.…”

Section: Autocorrelation Of Cyclizability Shows No Large-scale Patternssupporting

confidence: 87%

“…Dinucleotide composition is a driving factor in local DNA-rigidity Previous studies (29,30) as well as the model created in this study showed high accuracy of computationally predicting cyclizability values of the loop-seq data (r=0.94, Pearson correlation of actual vs. observed cyclizability), suggesting that there is a set of "rules" determining the rigidity of DNA that the model was able to learn and that rigidity is determined by sequence. Previous studies concluded that base composition is not a good indicator of DNA rigidity, except for mono-base or two-base repeats (44).…”

Section: Autocorrelation Of Cyclizability Shows No Large-scale Patternsmentioning

confidence: 66%

“…In this study, a deep learning CNN/LSTM model, CycPred, was developed and utilized to further explore the biological functions of local DNA flexibility. Our work builds on the previously published studies on experimentally measured (28) and computationally predicted DNA cyclizability scores (29,30). We investigated the significance of mechanical flexibility in several species from different kingdoms, including plants, with regard to magnitude relative to random expectation, underlying sequence determinants, sequence range, and genomic features, such as transcription start sites, TF-binding sites, single nucleotide polymorphisms (SNPs) and de novo mutations, DNA-methylation-, and crossing-over sites.…”

Section: Introductionmentioning

confidence: 93%

“…An accurate computational prediction model may find many applications and may even potentially replace experimental methods, while also leading to a higher base-pair resolution. Accordingly, attempts have been made and high prediction accuracies have been reported (29,30).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Predictions of DNA mechanical properties at a genomic scale reveal potentially new functional roles of DNA-flexibility

Back

Walther

2023

Preprint

View full text Add to dashboard Cite

Mechanical properties of DNA have been implied to influence many its biological functions. Recently, a new high-throughput method, called loop-seq, that allows measuring the intrinsic bendability of DNA fragments, has been developed. Using loop-seq data, we created a deep learning model to explore the biological significance of local DNA flexibility in a range of different species from different kingdoms. Consistently, we observed a characteristic and largely nucleotide-composition-driven change of local flexibility near transcription start sites. No evidence of a generally present region of lowered flexibility upstream of transcription start sites to facilitate transcription factor binding was found. Yet, depending on the actual transcription factor investigated, flanking-sequence-dependent DNA flexibility was identified as a potential factor influencing binding. Compared to randomized genomic sequences, depending on species and taxa, actual genomic sequences were observed both with increased and lowered flexibility. Furthermore, in Arabidopsis thaliana, crossing-over and mutation rates, both de novo and fixed, were found to be linked to rigid sequence regions. Our study presents a range of significant correlations between characteristic DNA mechanical properties and genomic features, the significance of which with regard to detailed molecular relevance awaits further experimental and theoretical exploration.

show abstract

Section: Autocorrelation Of Cyclizability Shows No Large-scale Patternssupporting

confidence: 93%

Section: Autocorrelation Of Cyclizability Shows No Large-scale Patternssupporting

confidence: 87%

Section: Autocorrelation Of Cyclizability Shows No Large-scale Patternsmentioning

confidence: 66%

Section: Introductionmentioning

confidence: 93%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Predictions of DNA mechanical properties at a genomic scale reveal potentially new functional roles of DNA-flexibility

Back

Walther

2023

Preprint

View full text Add to dashboard Cite

show abstract

Predicting DNA structure using a deep learning method

Li,

Chiu,

Rohs

2024

Nat Commun

View full text Add to dashboard Cite

Understanding the mechanisms of protein-DNA binding is critical in comprehending gene regulation. Three-dimensional DNA structure, also described as DNA shape, plays a key role in these mechanisms. In this study, we present a deep learning-based method, Deep DNAshape, that fundamentally changes the current k-mer based high-throughput prediction of DNA shape features by accurately accounting for the influence of extended flanking regions, without the need for extensive molecular simulations or structural biology experiments. By using the Deep DNAshape method, DNA structural features can be predicted for any length and number of DNA sequences in a high-throughput manner, providing an understanding of the effects of flanking regions on DNA structure in a target region of a sequence. The Deep DNAshape method provides access to the influence of distant flanking regions on a region of interest. Our findings reveal that DNA shape readout mechanisms of a core target are quantitatively affected by flanking regions, including extended flanking regions, providing valuable insights into the detailed structural readout mechanisms of protein-DNA binding. Furthermore, when incorporated in machine learning models, the features generated by Deep DNAshape improve the model prediction accuracy. Collectively, Deep DNAshape can serve as versatile and powerful tool for diverse DNA structure-related studies.

show abstract

Assessing base-resolution DNA mechanics on the genome scale

Jiang,

Hu,

Lai

et al. 2023

Nucleic Acids Research

View full text Add to dashboard Cite

Intrinsic DNA properties including bending play a crucial role in diverse biological systems. A recent advance in a high-throughput technology called loop-seq makes it possible to determine the bendability of hundred thousand 50-bp DNA duplexes in one experiment. However, it's still challenging to assess base-resolution sequence bendability in large genomes such as human, which requires thousands of such experiments. Here, we introduce ‘BendNet’—a deep neural network to predict the intrinsic DNA bending at base-resolution by using loop-seq results in yeast as training data. BendNet can predict the DNA bendability of any given sequence from different species with high accuracy. To explore the utility of BendNet, we applied it to the human genome and observed DNA bendability is associated with chromatin features and disease risk regions involving transcription/enhancer regulation, DNA replication, transcription factor binding and extrachromosomal circular DNA generation. These findings expand our understanding on DNA mechanics and its association with transcription regulation in mammals. Lastly, we built a comprehensive resource of genomic DNA bendability profiles for 307 species by applying BendNet, and provided an online tool to assess the bendability of user-specified DNA sequences (http://www.dnabendnet.com/).

show abstract

DeepBend: An interpretable model of DNA bendability

Cited by 6 publications

References 37 publications

Predictions of DNA mechanical properties at a genomic scale reveal potentially new functional roles of DNA-flexibility

Predictions of DNA mechanical properties at a genomic scale reveal potentially new functional roles of DNA-flexibility

Predicting DNA structure using a deep learning method

Assessing base-resolution DNA mechanics on the genome scale

Contact Info

Product

Resources

About