Graph Neural Networks for Metagenomic Binning

Andre Lamurias; Alessandro Tibo; Katja Hose; Mads Albertsen; Thomas Dyhre Nielsen

Graph Neural Networks for Metagenomic Binning

Andre Lamurias, Alessandro Tibo, Katja Hose, Mads Albertsen, Thomas Dyhre Nielsen

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

Abstract

Most methods for metagenomic binning rely solely on the local properties of the individual contigs. Because of this, these techniques are unable to take advantage of the connections between contigs as established by the assembly graph. In this paper, we explore Graph Neural Networks (GNNs) to leverage the assembly graph when learning contig representations for metagenomic binning. We applied four different types of GNN architectures, comparing their results on real and synthetic datasets, demonstrating encouraging results and, therefore, a promising research direction to pursue and explore.

Original language	English
Title of host publication	The 2023 ICML Workshop on Computational Biology
Number of pages	7
Publication date	2023
Article number	81
Publication status	Published - 2023
Event	The 2023 ICML Workshop on Computational Biology - Honolulu, United States Duration: 29 Jul 2023 → 29 Jul 2023 https://icml-compbio.github.io/index.html

Workshop

Workshop	The 2023 ICML Workshop on Computational Biology
Country/Territory	United States
City	Honolulu
Period	29/07/2023 → 29/07/2023
Internet address	https://icml-compbio.github.io/index.html

Access to Document

https://icml-compbio.github.io/2023/papers/WCBICML2023_paper81.pdf

AUB Link

Search for the material in Aalborg University Library's search engine

1 Finished
1 Active

DarkScience: Illuminating microbial dark matter through data science
Albertsen, M. (PI), Hose, K. (PI), Nielsen, T. D. (PI), Heidelbach, S. (Project Participant), Knudsen, K. S. (Project Participant), Sereika, M. (Project Participant), Corfixen, M. (Project Participant), Celikkanat, A. (Project Participant), Masegosa, A. (Project Participant), Sagi, T. (Project Participant), Nissen, J. (Project Participant), Heede, T. (Project Participant) & Kirkegaard, R. H. (Project Participant)
Villum Foundation
01/11/2022 → …
Project: Research
Data Science meets Microbial Dark Matter
Albertsen, M. (PI), Hose, K. (PI), Nielsen, T. D. (PI), Lamurias, A. (Project Participant) & Mølvang Dall, S. (Project Participant)
Danish E-infrastructure Cooperation, Villum Foundation
01/01/2021 → 31/12/2023
Project: Research

Cite this

@inproceedings{4297165538794e03aa3c172b9df16601,

title = "Graph Neural Networks for Metagenomic Binning",

abstract = "Most methods for metagenomic binning rely solely on the local properties of the individual contigs. Because of this, these techniques are unable to take advantage of the connections between contigs as established by the assembly graph. In this paper, we explore Graph Neural Networks (GNNs) to leverage the assembly graph when learning contig representations for metagenomic binning. We applied four different types of GNN architectures, comparing their results on real and synthetic datasets, demonstrating encouraging results and, therefore, a promising research direction to pursue and explore.",

author = "Andre Lamurias and Alessandro Tibo and Katja Hose and Mads Albertsen and Nielsen, {Thomas Dyhre}",

year = "2023",

language = "English",

booktitle = "The 2023 ICML Workshop on Computational Biology",

note = "The 2023 ICML Workshop on Computational Biology ; Conference date: 29-07-2023 Through 29-07-2023",

url = "https://icml-compbio.github.io/index.html",

}

TY - GEN

T1 - Graph Neural Networks for Metagenomic Binning

AU - Lamurias, Andre

AU - Tibo, Alessandro

AU - Hose, Katja

AU - Albertsen, Mads

AU - Nielsen, Thomas Dyhre

PY - 2023

Y1 - 2023

N2 - Most methods for metagenomic binning rely solely on the local properties of the individual contigs. Because of this, these techniques are unable to take advantage of the connections between contigs as established by the assembly graph. In this paper, we explore Graph Neural Networks (GNNs) to leverage the assembly graph when learning contig representations for metagenomic binning. We applied four different types of GNN architectures, comparing their results on real and synthetic datasets, demonstrating encouraging results and, therefore, a promising research direction to pursue and explore.

AB - Most methods for metagenomic binning rely solely on the local properties of the individual contigs. Because of this, these techniques are unable to take advantage of the connections between contigs as established by the assembly graph. In this paper, we explore Graph Neural Networks (GNNs) to leverage the assembly graph when learning contig representations for metagenomic binning. We applied four different types of GNN architectures, comparing their results on real and synthetic datasets, demonstrating encouraging results and, therefore, a promising research direction to pursue and explore.

M3 - Article in proceeding

BT - The 2023 ICML Workshop on Computational Biology

T2 - The 2023 ICML Workshop on Computational Biology

Y2 - 29 July 2023 through 29 July 2023

ER -

Graph Neural Networks for Metagenomic Binning

Abstract

Workshop

Access to Document

AUB Link

Fingerprint

Projects

DarkScience: Illuminating microbial dark matter through data science

Data Science meets Microbial Dark Matter

Cite this