EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Global-Frequency-Weighted Symmetrized KL Dissimilarity for Alignment-Free Genomic Sequence Comparison

Pegah Yarahmadi, Chiara Panico, Filippo Gandino, Renato Ferrero

BIO Web of Conferences · 2026 · 267 · 02001

DOI: 10.1051/bioconf/202626702001 ↗

Kurz gefasst

Fachartikel in der Zeitschrift BIO Web of Conferences, online erschienen am 18.11.2026. Verfasst von Pegah Yarahmadi und 3 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Alignment-free genomic sequence comparison commonly relies on k-mer profiles, but treating all k-mers equally may obscure informative compositional differences. This study evaluates global-frequency-weighted symmetrized Kullback–Leibler (KL) dissimilarities for k-mer probability distributions. Logarithmic, inverse-frequency, and inverse-square-root weighting are assessed together with symmetrized KL on normalized TF–IDF profiles and the baselines unweighted symmetrized KL dissimilarity, Jensen–Shannon distance, cosine distance, and KWIP. Experiments use a controlled ten-sequence dataset from the human ASH1L region and 23 genomes from non-singleton classes of a Brucella dataset. On ASH1L , all methods yield low silhouette scores, with cosine distance obtaining the highest value (0.0672). On Brucella , logarithmic weighting obtains the highest value (0.384), followed by inverse-square-root weighting (0.356). Moderate frequency weighting increases label-consistent separation in the whole-genome benchmark but provides no consistent advantage when the predefined groups are weakly separated.

Zitieren

APA

Yarahmadi, P., Panico, C., Gandino, F., & Ferrero, R. (2026). Global-Frequency-Weighted Symmetrized KL Dissimilarity for Alignment-Free Genomic Sequence Comparison. BIO Web of Conferences, 267, 02001. https://doi.org/10.1051/bioconf/202626702001

Vancouver

Yarahmadi P, Panico C, Gandino F, Ferrero R. Global-Frequency-Weighted Symmetrized KL Dissimilarity for Alignment-Free Genomic Sequence Comparison. BIO Web of Conferences. 2026;267:02001. doi:10.1051/bioconf/202626702001

BibTeX
@article{Yarahmadi2026,
  author = {Yarahmadi, Pegah and Panico, Chiara and Gandino, Filippo and Ferrero, Renato},
  title = {Global-Frequency-Weighted Symmetrized KL Dissimilarity for Alignment-Free Genomic Sequence Comparison},
  journal = {BIO Web of Conferences},
  year = {2026},
  volume = {267},
  pages = {02001},
  doi = {10.1051/bioconf/202626702001}
}
Kostenlos starten →