miRBench: novel benchmark datasets for microRNA binding site
prediction that mitigate against prevalent microRNA frequency
class bias

J 2025

miRBench: novel benchmark datasets for microRNA binding site prediction that mitigate against prevalent microRNA frequency class bias

SAMMUT, Stephanie; Katarína GREŠOVÁ; Dimosthenis TZIMOTOUDIS; Eva MARŠÁLKOVÁ; David ČECHÁK et al.

Základní údaje

Originální název

miRBench: novel benchmark datasets for microRNA binding site prediction that mitigate against prevalent microRNA frequency class bias

Autoři

SAMMUT, Stephanie; Katarína GREŠOVÁ; Dimosthenis TZIMOTOUDIS; Eva MARŠÁLKOVÁ ; David ČECHÁK a Panagiotis ALEXIOU

Vydání

Computational systems bioinformatics / Life Sciences Society. Computational Systems Bioinformatics Conference, OXFORD UNIV PRESS, 2025, 1367-4803

Další údaje

Jazyk

angličtina

Typ výsledku

Článek v odborném periodiku

Obor

10609 Biochemical research methods

Stát vydavatele

Velká Británie a Severní Irsko

Utajení

není předmětem státního či obchodního tajemství

Odkazy

URL

Impakt faktor

Impact factor: 5.400 v roce 2024

Označené pro přenos do RIV

Ano

Kód RIV

RIV/00216224:14740/25:00141656

Organizační jednotka

Středoevropský technologický institut

Klíčová slova anglicky

miRNA; THERAPEUTICS; RNAS

Štítky

143160, CF BIOIT, CF GEN, rivok

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 14. 3. 2026 21:22, Mgr. Eva Dubská

Anotace

V originále

MicroRNAs (miRNAs) are crucial regulators of gene expression, but the precise mechanisms governing their binding to target sites remain unclear. A major contributing factor to this is the lack of unbiased experimental datasets for training accurate prediction models. While recent experimental advances have provided numerous miRNA–target interactions, these are solely positive interactions. Generating negative examples in silico is challenging and prone to introducing biases, such as the miRNA frequency class bias identified in this work. Biases within datasets can compromise model generalization, leading models to learn dataset-specific artifacts rather than true biological patterns.

Návaznosti

90254, velká výzkumná infrastruktura

Název: e-INFRA CZ II

90267, velká výzkumná infrastruktura

Název: NCMG III

Citovat

SAMMUT, Stephanie; Katarína GREŠOVÁ; Dimosthenis TZIMOTOUDIS; Eva MARŠÁLKOVÁ; David ČECHÁK a Panagiotis ALEXIOU. miRBench: novel benchmark datasets for microRNA binding site prediction that mitigate against prevalent microRNA frequency class bias. Computational systems bioinformatics / Life Sciences Society. Computational Systems Bioinformatics Conference. OXFORD UNIV PRESS, 2025, roč. 41, č. 1, s. 542-551, 11 s. ISSN 1367-4803. Dostupné z: https://doi.org/10.1093/bioinformatics/btaf233.

@article{2509537,
   author = {Sammut, Stephanie and Grešová, Katarína and Tzimotoudis, Dimosthenis and Maršálková, Eva and Čechák, David and Alexiou, Panagiotis},
   article_number = {1},
   doi = {https://doi.org/10.1093/bioinformatics/btaf233},
   keywords = {miRNA; THERAPEUTICS; RNAS},
   language = {eng},
   issn = {1367-4803},
   journal = {Computational systems bioinformatics / Life Sciences Society. Computational Systems Bioinformatics Conference},
   title = {miRBench: novel benchmark datasets for microRNA binding site prediction that mitigate against prevalent microRNA frequency class bias},
   url = {https://academic.oup.com/bioinformatics/article/41/Supplement_1/i542/8199406?login=true},
   volume = {41},
   year = {2025}
}

TY  - JOUR
ID  - 2509537
AU  - Sammut, Stephanie - Grešová, Katarína - Tzimotoudis, Dimosthenis - Maršálková, Eva - Čechák, David - Alexiou, Panagiotis
PY  - 2025
TI  - miRBench: novel benchmark datasets for microRNA binding site prediction that mitigate against prevalent microRNA frequency class bias
JF  - Computational systems bioinformatics / Life Sciences Society. Computational Systems Bioinformatics Conference
VL  - 41
IS  - 1
SP  - 542-551
EP  - 542-551
PB  - OXFORD UNIV PRESS
SN  - 13674803
KW  - miRNA
KW  - THERAPEUTICS
KW  - RNAS
UR  - https://academic.oup.com/bioinformatics/article/41/Supplement_1/i542/8199406?login=true
N2  - MicroRNAs (miRNAs) are crucial regulators of gene expression, but the precise mechanisms governing their binding to target sites remain unclear. A major contributing factor to this is the lack of unbiased experimental datasets for training accurate prediction models. While recent experimental advances have provided numerous miRNA–target interactions, these are solely positive interactions. Generating negative examples in silico is challenging and prone to introducing biases, such as the miRNA frequency class bias identified in this work. Biases within datasets can compromise model generalization, leading models to learn dataset-specific artifacts rather than true biological patterns.
ER  -

SAMMUT, Stephanie; Katarína GREŠOVÁ; Dimosthenis TZIMOTOUDIS; Eva MARŠÁLKOVÁ; David ČECHÁK a Panagiotis ALEXIOU. miRBench: novel benchmark datasets for microRNA binding site prediction that mitigate against prevalent microRNA frequency class bias. \textit{Computational systems bioinformatics / Life Sciences Society. Computational Systems Bioinformatics Conference}. OXFORD UNIV PRESS, 2025, roč.~41, č.~1, s.~542-551, 11 s. ISSN~1367-4803. Dostupné z: https://doi.org/10.1093/bioinformatics/btaf233.

Přehled o publikaci