Dataset Information

Deep statistical modelling of nanopore sequencing translocation times reveals latent non-B DNA structures.

ABSTRACT:

Motivation

Non-canonical (or non-B) DNA are genomic regions whose three-dimensional conformation deviates from the canonical double helix. Non-B DNA play an important role in basic cellular processes and are associated with genomic instability, gene regulation, and oncogenesis. Experimental methods are low-throughput and can detect only a limited set of non-B DNA structures, while computational methods rely on non-B DNA base motifs, which are necessary but not sufficient indicators of non-B structures. Oxford Nanopore sequencing is an efficient and low-cost platform, but it is currently unknown whether nanopore reads can be used for identifying non-B structures.

Results

We build the first computational pipeline to predict non-B DNA structures from nanopore sequencing. We formalize non-B detection as a novelty detection problem and develop the GoFAE-DND, an autoencoder that uses goodness-of-fit (GoF) tests as a regularizer. A discriminative loss encourages non-B DNA to be poorly reconstructed and optimizing Gaussian GoF tests allows for the computation of P-values that indicate non-B structures. Based on whole genome nanopore sequencing of NA12878, we show that there exist significant differences between the timing of DNA translocation for non-B DNA bases compared with B-DNA. We demonstrate the efficacy of our approach through comparisons with novelty detection methods using experimental data and data synthesized from a new translocation time simulator. Experimental validations suggest that reliable detection of non-B DNA from nanopore sequencing is achievable.

Availability and implementation

Source code is available at https://github.com/bayesomicslab/ONT-nonb-GoFAE-DND.

SUBMITTER: Hosseini M

PROVIDER: S-EPMC10311326 | biostudies-literature | 2023 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep statistical modelling of nanopore sequencing translocation times reveals latent non-B DNA structures.

Hosseini Marjan M Palmer Aaron A Manka William W Grady Patrick G S PGS Patchigolla Venkata V Bi Jinbo J O'Neill Rachel J RJ Chi Zhiyi Z Aguiar Derek D

Bioinformatics (Oxford, England) 20230601 39 Suppl 1

<h4>Motivation</h4>Non-canonical (or non-B) DNA are genomic regions whose three-dimensional conformation deviates from the canonical double helix. Non-B DNA play an important role in basic cellular processes and are associated with genomic instability, gene regulation, and oncogenesis. Experimental methods are low-throughput and can detect only a limited set of non-B DNA structures, while computational methods rely on non-B DNA base motifs, which are necessary but not sufficient indicators of no ...[more]

PMID: 37387144

Similar Datasets

Project description:Viroids are small circular single-stranded infectious RNAs characterized by a relatively high mutation level. Knowledge of their sequence heterogeneity remains largely elusive and previous studies, using Sanger sequencing, were based on a limited number of sequences. In an attempt to address sequence heterogeneity from a population dynamics perspective, a GF305-indicator peach tree was infected with a single variant of the Avsunviroidae family member Peach latent mosaic viroid (PLMVd). Six months post-inoculation, full-length circular conformers of PLMVd were isolated and deep-sequenced. We devised an original approach to the bioinformatics refinement of our sequence libraries involving important phenotypic data, based on the systematic analysis of hammerhead self-cleavage activity. Two distinct libraries yielded a total of 3,939 different PLMVd variants. Sequence variants exhibiting up to ∼17% of mutations relative to the inoculated viroid were retrieved, clearly illustrating the high level of divergence dynamics within a unique population. While we initially assumed that most positions of the viroid sequence would mutate, we were surprised to discover that ∼50% of positions remained perfectly conserved, including several small stretches as well as a small motif reminiscent of a GNRA tetraloop which are the result of various selective pressures. Using a hierarchical clustering algorithm, the different variants harvested were subdivided into 7 clusters. We found that most sequences contained an average of 4.6 to 6.4 mutations compared to the variant used to initially inoculate the plant. Interestingly, it was possible to reconstitute and compare the sequence evolution of each of these clusters. In doing so, we identified several key mutations. This study provides a reliable pipeline for the treatment of viroid deep-sequencing. It also sheds new light on the extent of sequence variation that a viroid population can sustain, and which may give rise to a quasispecies.

Project description:BackgroundThe advancements in DNA sequencing technologies have allowed researchers to progress from the analyses of a single organism towards the deep sequencing of a sample of organisms. With sufficient sequencing depth, it is now possible to detect subtle variations between members of the same species, or between mixed species with shared biomarkers, such as the 16S rRNA gene. However, traditional sequencing analyses of samples from largely homogeneous populations are often still based on multiple sequence alignments (MSA), where each sequence is placed along a separate row and similarities between aligned bases can be followed down each column. While this visual format is intuitive for a small set of aligned sequences, the representation quickly becomes cumbersome as sequencing depths cover loci hundreds or thousands of reads deep.FindingsWe have developed ANDES, a software library and a suite of applications, written in Perl and R, for the statistical ANalyses of DEep Sequencing. The fundamental data structure underlying ANDES is the position profile, which contains the nucleotide distributions for each genomic position resultant from a multiple sequence alignment (MSA). Tools include the root mean square deviation (RMSD) plot, which allows for the visual comparison of multiple samples on a position-by-position basis, and the computation of base conversion frequencies (transition/transversion rates), variation (Shannon entropy), inter-sample clustering and visualization (dendrogram and multidimensional scaling (MDS) plot), threshold-driven consensus sequence generation and polymorphism detection, and the estimation of empirically determined sequencing quality values.ConclusionsAs new sequencing technologies evolve, deep sequencing will become increasingly cost-efficient and the inter and intra-sample comparisons of largely homogeneous sequences will become more common. We have provided a software package and demonstrated its application on various empirically-derived datasets. Investigators may download the software from Sourceforge at https://sourceforge.net/projects/andestools.

Dataset Information

Deep statistical modelling of nanopore sequencing translocation times reveals latent non-B DNA structures.

Motivation

Results

Availability and implementation

Publications

Deep statistical modelling of nanopore sequencing translocation times reveals latent non-B DNA structures.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets