Transfer Learning Allows Accurate RBP Target Site Prediction
with Limited Sample Sizes

J 2023

Transfer Learning Allows Accurate RBP Target Site Prediction with Limited Sample Sizes

VACULÍK, Ondřej; Eliška CHALUPOVÁ; Katarína GREŠOVÁ; Tomáš MAJTNER; Panagiotis ALEXIOU et al.

Základní údaje

Originální název

Transfer Learning Allows Accurate RBP Target Site Prediction with Limited Sample Sizes

Autoři

VACULÍK, Ondřej; Eliška CHALUPOVÁ; Katarína GREŠOVÁ; Tomáš MAJTNER a Panagiotis ALEXIOU

Vydání

Biology, MDPI, 2023, 2079-7737

Další údaje

Jazyk

angličtina

Typ výsledku

Článek v odborném periodiku

Obor

10201 Computer sciences, information science, bioinformatics

Stát vydavatele

Švýcarsko

Utajení

není předmětem státního či obchodního tajemství

Odkazy

URL

Impakt faktor

Impact factor: 3.600

Označené pro přenos do RIV

Ano

Kód RIV

RIV/00216224:14310/23:00131807

Organizační jednotka

Přírodovědecká fakulta

Klíčová slova anglicky

RNA-binding protein; CLIP-seq; deep learning; transfer learning; interpretation

Štítky

143160, CF BIOIT, rivok

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 9. 10. 2024 13:26, Mgr. Eva Dubská

Anotace

V originále

RNA-binding proteins are vital regulators in numerous biological processes. Their disfunction can result in diverse diseases, such as cancer or neurodegenerative disorders, making the prediction of their binding sites of high importance. Deep learning (DL) has brought about a revolution in various biological domains, including the field of protein–RNA interactions. Nonetheless, several challenges persist, such as the limited availability of experimentally validated binding sites to train well-performing DL models for the majority of proteins. Here, we present a novel training approach based on transfer learning (TL) to address the issue of limited data. Employing a sophisticated and interpretable architecture, we compare the performance of our method trained using two distinct approaches: training from scratch (SCR) and utilizing TL. Additionally, we benchmark our results against the current state-of-the-art methods. Furthermore, we tackle the challenges associated with selecting appropriate input features and determining optimal interval sizes. Our results show that TL enhances model performance, particularly in datasets with minimal training data, where satisfactory results can be achieved with just a few hundred RNA binding sites. Moreover, we demonstrate that integrating both sequence and evolutionary conservation information leads to superior performance. Additionally, we showcase how incorporating an attention layer into the model facilitates the interpretation of predictions within a biologically relevant context.

Návaznosti

EF18_053/0016952, projekt VaV

Název: Postdoc2MUNI

90267, velká výzkumná infrastruktura

Název: NCMG III

Citovat

VACULÍK, Ondřej; Eliška CHALUPOVÁ; Katarína GREŠOVÁ; Tomáš MAJTNER a Panagiotis ALEXIOU. Transfer Learning Allows Accurate RBP Target Site Prediction with Limited Sample Sizes. Biology. MDPI, 2023, roč. 12, č. 10, s. 1-19. ISSN 2079-7737. Dostupné z: https://doi.org/10.3390/biology12101276.

@article{2319117,
   author = {Vaculík, Ondřej and Chalupová, Eliška and Grešová, Katarína and Majtner, Tomáš and Alexiou, Panagiotis},
   article_number = {10},
   doi = {https://doi.org/10.3390/biology12101276},
   keywords = {RNA-binding protein; CLIP-seq; deep learning; transfer learning; interpretation},
   language = {eng},
   issn = {2079-7737},
   journal = {Biology},
   title = {Transfer Learning Allows Accurate RBP Target Site Prediction with Limited Sample Sizes},
   url = {https://www.mdpi.com/2079-7737/12/10/1276},
   volume = {12},
   year = {2023}
}

TY  - JOUR
ID  - 2319117
AU  - Vaculík, Ondřej - Chalupová, Eliška - Grešová, Katarína - Majtner, Tomáš - Alexiou, Panagiotis
PY  - 2023
TI  - Transfer Learning Allows Accurate RBP Target Site Prediction with Limited Sample Sizes
JF  - Biology
VL  - 12
IS  - 10
SP  - 1-19
EP  - 1-19
PB  - MDPI
SN  - 20797737
KW  - RNA-binding protein
KW  - CLIP-seq
KW  - deep learning
KW  - transfer learning
KW  - interpretation
UR  - https://www.mdpi.com/2079-7737/12/10/1276
N2  - RNA-binding proteins are vital regulators in numerous biological processes. Their disfunction can result in diverse diseases, such as cancer or neurodegenerative disorders, making the prediction of their binding sites of high importance. Deep learning (DL) has brought about a revolution in various biological domains, including the field of protein–RNA interactions. Nonetheless, several challenges persist, such as the limited availability of experimentally validated binding sites to train well-performing DL models for the majority of proteins. Here, we present a novel training approach based on transfer learning (TL) to address the issue of limited data. Employing a sophisticated and interpretable architecture, we compare the performance of our method trained using two distinct approaches: training from scratch (SCR) and utilizing TL. Additionally, we benchmark our results against the current state-of-the-art methods. Furthermore, we tackle the challenges associated with selecting appropriate input features and determining optimal interval sizes. Our results show that TL enhances model performance, particularly in datasets with minimal training data, where satisfactory results can be achieved with just a few hundred RNA binding sites. Moreover, we demonstrate that integrating both sequence and evolutionary conservation information leads to superior performance. Additionally, we showcase how incorporating an attention layer into the model facilitates the interpretation of predictions within a biologically relevant context.
ER  -

VACULÍK, Ondřej; Eliška CHALUPOVÁ; Katarína GREŠOVÁ; Tomáš MAJTNER a Panagiotis ALEXIOU. Transfer Learning Allows Accurate RBP Target Site Prediction with Limited Sample Sizes. \textit{Biology}. MDPI, 2023, roč.~12, č.~10, s.~1-19. ISSN~2079-7737. Dostupné z: https://doi.org/10.3390/biology12101276.

Přehled o publikaci