Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat
Sign Up
Cited by 7Fractal and DNA sequence analysisDNA and Nucleic Acid ChemistryRNA and protein synthesis mechanisms

The DOI matches its Crossref record; 48 of the 48 references it cites were found in a registry or an index.

Quantum mechanical electronic and geometric parameters for DNA k-mers as features for machine learning

Open access
Masuda, Kairi, Abdullah, Adib A., Pflughaupt, Patrick, Sahakyan, Aleksandr B.·2024·Scientific Data
DOI 10.1038/s41597-024-03772-5PMC PMC11341866

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

We are witnessing a steep increase in model development initiatives in genomics that employ high-end machine learning methodologies. Of particular interest are models that predict certain genomic characteristics based solely on DNA sequence. These models, however, treat the DNA as a mere collection of four, A, T, G and C, letters, dismissing the past advancements in science that can enable the use of more intricate information from nucleic acid sequences. Here, we provide a comprehensive database of quantum mechanical (QM) and geometric features for all the permutations of 7-meric DNA in their representative B, A and Z conformations. The database is generated by employing the applicable high-cost and time-consuming QM methodologies. This can thus make it seamless to associate a wealth of novel molecular features to any DNA sequence, by scanning it with a matching k-meric window and pulling the pre-computed values from our database for further use in modelling. We demonstrate the usefulness of our deposited features through their exclusive use in developing a model for A->C mutation rates.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41597-024-03772-5

Cites

48 of 48 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Representative molecular structures of the 7-mer double-stranded DNA models used in this study. The structures are brought for (a) B-DNA, (b) A-DNA, and (c) Z-DNA conformations. For clarity, backbones are represented by ribbons. These structures have seven nucleotide spans at both strands.
Fig. 1 Representative molecular structures of the 7-mer double-stranded DNA models used in this study. The structures are brought for (a) B-DNA, (b) A-DNA, and (c) Z-DNA conformations. For clarity, backbones are represented by ribbons. These structures have seven nucleotide spans at both strands.CC BY · PubMed Central
Fig. 2 A schematic workflow of the dataset generation used in this study. In the whole procedure, we used R programming language for processing the data. (a) To make molecular DNA models, we used web 3DNA and NAB programmes, and a typical resultant DNA model is shown as an illustration. (b) Next, we optimised these models in two steps. First via MM optimisation by AmberTools21, next via QM optimisation by MOPAC2016. (c1) After optimisation, we extracted electronic features by single-point MOPAC2016 calculations, and geometric features by using Curves+. For charges and populations, we calculated the maximum, minimum, and mean values of phosphate, sugar, and base moieties. (c2) We further conducted single-point calculations for each strand separately, and for the states with deleted central base. Next, the differences in energy and overall features between the duplex and these states were calculated. Note, that differences in IP, HOMO, and LUMO (ΔIP, ΔHOMO, and ΔLUMO) from varying molecular states do not have clear physical meaning and should be avoided from direct usage in machine learning. However, we include them in our dataset to enable the retrieval of the corresponding values for single-stranded and base-deleted states from the IP, HOMO, and LUMO of duplex states, i.e. by taking IP-ΔIP, HOMO-ΔHOMO and LUMO-ΔLUMO.
Fig. 2 A schematic workflow of the dataset generation used in this study. In the whole procedure, we used R programming language for processing the data. (a) To make molecular DNA models, we used web 3DNA and NAB programmes, and a typical resultant DNA model is shown as an illustration. (b) Next, we optimised these models in two steps. First via MM optimisation by AmberTools21, next via QM optimisation by MOPAC2016. (c1) After optimisation, we extracted electronic features by single-point MOPAC2016 calculations, and geometric features by using Curves+. For charges and populations, we calculated the maximum, minimum, and mean values of phosphate, sugar, and base moieties. (c2) We further conducted single-point calculations for each strand separately, and for the states with deleted central base. Next, the differences in energy and overall features between the duplex and these states were calculated. Note, that differences in IP, HOMO, and LUMO (ΔIP, ΔHOMO, and ΔLUMO) from varying molecular states do not have clear physical meaning and should be avoided from direct usage in machine learning. However, we include them in our dataset to enable the retrieval of the corresponding values for single-stranded and base-deleted states from the IP, HOMO, and LUMO of duplex states, i.e. by taking IP-ΔIP, HOMO-ΔHOMO and LUMO-ΔLUMO.CC BY · PubMed Central
Fig. 3 Developing a sequence-driven machine learning model for A to C mutation rates based solely on QM features mapped onto the sequence. Schematic illustrations are shown for (a) combining our QM dataset and an external dataset for A to C mutation rate constants (kA→C) and (b) constructing a machine learning model that predicts kA→C from the sequence, based on the sequence-mapped QM features deposited in the DNAkmerQM dataset presented in this work. (c) The mutation rate constants predicted through our machine learning model vs. actual mutation rate constants (in byr−1, i.e. mutation per billion years for each such nucleotide site). The dashed line shows the diagonal for the ideal match.
Fig. 3 Developing a sequence-driven machine learning model for A to C mutation rates based solely on QM features mapped onto the sequence. Schematic illustrations are shown for (a) combining our QM dataset and an external dataset for A to C mutation rate constants (kA→C) and (b) constructing a machine learning model that predicts kA→C from the sequence, based on the sequence-mapped QM features deposited in the DNAkmerQM dataset presented in this work. (c) The mutation rate constants predicted through our machine learning model vs. actual mutation rate constants (in byr−1, i.e. mutation per billion years for each such nucleotide site). The dashed line shows the diagonal for the ideal match.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Research works printed but not linked: 1

  1. Russo, N., Toscano, M. & Grand, A. Theoretical determination of electron affinity and ionization potential of DNA and RNA bases. Journal of Computational Chemistry 21, 1243–1250, https://doi.org/10.1002/1096-987X(20001115)21:14 (2000). · doi:10.1002/1096-987x(20001115)21:14

Metadata

Authors
Masuda, K., Abdullah, A. A., Pflughaupt, P., & Sahakyan, A. B.
Affiliations
University of Oxford (GB)
Corresponding author
Sahakyan, Aleksandr B.
ORCID
Venue
Scientific Data
Publisher
Springer Science and Business Media LLC
Published
2024-08-22
Volume
11
Issue
1
Pages
911-911
ISSN
2052-4463 (ISSN-L)
DOI
10.1038/s41597-024-03772-5
PMC
PMC11341866
OpenAlex
W4401814295
Cited by
7
References
48
Citations by year
2026: 4 · 2025: 1 · 2024: 2
Type
data paper
Language
en
Open access
Licence
Locations
  • Scientific Data (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Fractal and DNA sequence analysis
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › DNA and Nucleic Acid Chemistry
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › RNA and protein synthesis mechanisms
Keywords
geometric features, DNA sequence features, B-DNA conformation, feature database
MeSH terms
Machine Learning, DNA, Quantum Theory

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.