Details
Originalsprache | Englisch |
---|---|
Aufsatznummer | 572 |
Seitenumfang | 24 |
Fachzeitschrift | BMC BIOINFORMATICS |
Jahrgang | 22 |
Ausgabenummer | 1 |
Publikationsstatus | Veröffentlicht - 27 Nov. 2021 |
Abstract
Background: Viral infections are causing significant morbidity and mortality worldwide. Understanding the interaction patterns between a particular virus and human proteins plays a crucial role in unveiling the underlying mechanism of viral infection and pathogenesis. This could further help in prevention and treatment of virus-related diseases. However, the task of predicting protein–protein interactions between a new virus and human cells is extremely challenging due to scarce data on virus-human interactions and fast mutation rates of most viruses. Results: We developed a multitask transfer learning approach that exploits the information of around 24 million protein sequences and the interaction patterns from the human interactome to counter the problem of small training datasets. Instead of using hand-crafted protein features, we utilize statistically rich protein representations learned by a deep language modeling approach from a massive source of protein sequences. Additionally, we employ an additional objective which aims to maximize the probability of observing human protein–protein interactions. This additional task objective acts as a regularizer and also allows to incorporate domain knowledge to inform the virus-human protein–protein interaction prediction model. Conclusions: Our approach achieved competitive results on 13 benchmark datasets and the case study for the SARS-CoV-2 virus receptor. Experimental results show that our proposed model works effectively for both virus-human and bacteria-human protein–protein interaction prediction tasks. We share our code for reproducibility and future research at https://git.l3s.uni-hannover.de/dong/multitask-transfer.
ASJC Scopus Sachgebiete
- Biochemie, Genetik und Molekularbiologie (insg.)
- Strukturelle Biologie
- Biochemie, Genetik und Molekularbiologie (insg.)
- Biochemie
- Biochemie, Genetik und Molekularbiologie (insg.)
- Molekularbiologie
- Informatik (insg.)
- Angewandte Informatik
- Mathematik (insg.)
- Angewandte Mathematik
Ziele für nachhaltige Entwicklung
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
in: BMC BIOINFORMATICS, Jahrgang 22, Nr. 1, 572, 27.11.2021.
Publikation: Beitrag in Fachzeitschrift › Artikel › Forschung › Peer-Review
}
TY - JOUR
T1 - A multitask transfer learning framework for the prediction of virus-human protein–protein interactions
AU - Dong, Ngan Thi
AU - Brogden, Graham
AU - Gerold, Gisa
AU - Khosla, Megha
N1 - Funding Information: Open Access funding enabled and organized by Projekt DEAL. N.D is funded by VolkswagenStiftung’s initiative “Niedersächsisches Vorab” (Grant No.11-76251-99-3/19 (ZN3434)). G.B and G.G are supported by the Ministry of Lower Saxony (MWK, Project 76251-99 awarded to G.G.). M.K is supported the Federal Ministry of Education and Research (BMBF), Germany under the project LeibnizKILabor (Grant No. 01DD20003). The funding bodies did not play any role in the design of the study, collection, analysis, interpretation of data, and in writing the manuscript.
PY - 2021/11/27
Y1 - 2021/11/27
N2 - Background: Viral infections are causing significant morbidity and mortality worldwide. Understanding the interaction patterns between a particular virus and human proteins plays a crucial role in unveiling the underlying mechanism of viral infection and pathogenesis. This could further help in prevention and treatment of virus-related diseases. However, the task of predicting protein–protein interactions between a new virus and human cells is extremely challenging due to scarce data on virus-human interactions and fast mutation rates of most viruses. Results: We developed a multitask transfer learning approach that exploits the information of around 24 million protein sequences and the interaction patterns from the human interactome to counter the problem of small training datasets. Instead of using hand-crafted protein features, we utilize statistically rich protein representations learned by a deep language modeling approach from a massive source of protein sequences. Additionally, we employ an additional objective which aims to maximize the probability of observing human protein–protein interactions. This additional task objective acts as a regularizer and also allows to incorporate domain knowledge to inform the virus-human protein–protein interaction prediction model. Conclusions: Our approach achieved competitive results on 13 benchmark datasets and the case study for the SARS-CoV-2 virus receptor. Experimental results show that our proposed model works effectively for both virus-human and bacteria-human protein–protein interaction prediction tasks. We share our code for reproducibility and future research at https://git.l3s.uni-hannover.de/dong/multitask-transfer.
AB - Background: Viral infections are causing significant morbidity and mortality worldwide. Understanding the interaction patterns between a particular virus and human proteins plays a crucial role in unveiling the underlying mechanism of viral infection and pathogenesis. This could further help in prevention and treatment of virus-related diseases. However, the task of predicting protein–protein interactions between a new virus and human cells is extremely challenging due to scarce data on virus-human interactions and fast mutation rates of most viruses. Results: We developed a multitask transfer learning approach that exploits the information of around 24 million protein sequences and the interaction patterns from the human interactome to counter the problem of small training datasets. Instead of using hand-crafted protein features, we utilize statistically rich protein representations learned by a deep language modeling approach from a massive source of protein sequences. Additionally, we employ an additional objective which aims to maximize the probability of observing human protein–protein interactions. This additional task objective acts as a regularizer and also allows to incorporate domain knowledge to inform the virus-human protein–protein interaction prediction model. Conclusions: Our approach achieved competitive results on 13 benchmark datasets and the case study for the SARS-CoV-2 virus receptor. Experimental results show that our proposed model works effectively for both virus-human and bacteria-human protein–protein interaction prediction tasks. We share our code for reproducibility and future research at https://git.l3s.uni-hannover.de/dong/multitask-transfer.
KW - Human PPI
KW - Multitask
KW - Protein embedding
KW - Protein–protein interaction
KW - Transfer learning
KW - Virus-human PPI
UR - http://www.scopus.com/inward/record.url?scp=85120050210&partnerID=8YFLogxK
U2 - 10.1186/s12859-021-04484-y
DO - 10.1186/s12859-021-04484-y
M3 - Article
C2 - 34837942
AN - SCOPUS:85120050210
VL - 22
JO - BMC BIOINFORMATICS
JF - BMC BIOINFORMATICS
SN - 1471-2105
IS - 1
M1 - 572
ER -