Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Interpreting biologically informed neural networks for enhanced proteomic biomarker discovery and pathway analysis

Hartman, Erik ; Scott, Aaron M LU ; Karlsson, Christofer LU ; Mohanty, Tirthankar LU ; Vaara, Suvi T ; Linder, Adam LU ; Malmström, Lars LU and Malmström, Johan LU orcid (2023) In Nature Communications 14. p.1-13
Abstract

The incorporation of machine learning methods into proteomics workflows improves the identification of disease-relevant biomarkers and biological pathways. However, machine learning models, such as deep neural networks, typically suffer from lack of interpretability. Here, we present a deep learning approach to combine biological pathway analysis and biomarker identification to increase the interpretability of proteomics experiments. Our approach integrates a priori knowledge of the relationships between proteins and biological pathways and biological processes into sparse neural networks to create biologically informed neural networks. We employ these networks to differentiate between clinical subphenotypes of septic acute kidney... (More)

The incorporation of machine learning methods into proteomics workflows improves the identification of disease-relevant biomarkers and biological pathways. However, machine learning models, such as deep neural networks, typically suffer from lack of interpretability. Here, we present a deep learning approach to combine biological pathway analysis and biomarker identification to increase the interpretability of proteomics experiments. Our approach integrates a priori knowledge of the relationships between proteins and biological pathways and biological processes into sparse neural networks to create biologically informed neural networks. We employ these networks to differentiate between clinical subphenotypes of septic acute kidney injury and COVID-19, as well as acute respiratory distress syndrome of different aetiologies. To gain biological insight into the complex syndromes, we utilize feature attribution-methods to introspect the networks for the identification of proteins and pathways important for distinguishing between subtypes. The algorithms are implemented in a freely available open source Python-package ( https://github.com/InfectionMedicineProteomics/BINN ).

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Humans, COVID-19, Proteomics, Neural Networks, Computer, Algorithms, Acute Kidney Injury
in
Nature Communications
volume
14
article number
5359
pages
1 - 13
publisher
Nature Publishing Group
external identifiers
  • scopus:85169531696
  • pmid:37660105
ISSN
2041-1723
DOI
10.1038/s41467-023-41146-4
language
English
LU publication?
yes
additional info
© 2023. Springer Nature Limited.
id
58c7100c-ee72-4ea3-a4e4-8d0e9191fbd7
date added to LUP
2023-09-12 08:21:29
date last changed
2024-04-20 03:05:44
@article{58c7100c-ee72-4ea3-a4e4-8d0e9191fbd7,
  abstract     = {{<p>The incorporation of machine learning methods into proteomics workflows improves the identification of disease-relevant biomarkers and biological pathways. However, machine learning models, such as deep neural networks, typically suffer from lack of interpretability. Here, we present a deep learning approach to combine biological pathway analysis and biomarker identification to increase the interpretability of proteomics experiments. Our approach integrates a priori knowledge of the relationships between proteins and biological pathways and biological processes into sparse neural networks to create biologically informed neural networks. We employ these networks to differentiate between clinical subphenotypes of septic acute kidney injury and COVID-19, as well as acute respiratory distress syndrome of different aetiologies. To gain biological insight into the complex syndromes, we utilize feature attribution-methods to introspect the networks for the identification of proteins and pathways important for distinguishing between subtypes. The algorithms are implemented in a freely available open source Python-package ( https://github.com/InfectionMedicineProteomics/BINN ).</p>}},
  author       = {{Hartman, Erik and Scott, Aaron M and Karlsson, Christofer and Mohanty, Tirthankar and Vaara, Suvi T and Linder, Adam and Malmström, Lars and Malmström, Johan}},
  issn         = {{2041-1723}},
  keywords     = {{Humans; COVID-19; Proteomics; Neural Networks, Computer; Algorithms; Acute Kidney Injury}},
  language     = {{eng}},
  pages        = {{1--13}},
  publisher    = {{Nature Publishing Group}},
  series       = {{Nature Communications}},
  title        = {{Interpreting biologically informed neural networks for enhanced proteomic biomarker discovery and pathway analysis}},
  url          = {{http://dx.doi.org/10.1038/s41467-023-41146-4}},
  doi          = {{10.1038/s41467-023-41146-4}},
  volume       = {{14}},
  year         = {{2023}},
}