Protocol to explain graph neural network predictions using an edge-centric Shapley value-based approach
Summary: Here we present EdgeSHAPer, a workflow for explaining graph neural networks by approximating Shapley values using Monte Carlo sampling. In this protocol, we describe steps to execute Python scripts for a chemical dataset from the original publication; however, this approach is also applicab...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Elsevier
2022-12-01
|
Series: | STAR Protocols |
Subjects: | |
Online Access: | http://www.sciencedirect.com/science/article/pii/S2666166722007675 |
_version_ | 1797980340760870912 |
---|---|
author | Andrea Mastropietro Giuseppe Pasculli Jürgen Bajorath |
author_facet | Andrea Mastropietro Giuseppe Pasculli Jürgen Bajorath |
author_sort | Andrea Mastropietro |
collection | DOAJ |
description | Summary: Here we present EdgeSHAPer, a workflow for explaining graph neural networks by approximating Shapley values using Monte Carlo sampling. In this protocol, we describe steps to execute Python scripts for a chemical dataset from the original publication; however, this approach is also applicable to any user-provided dataset. We also detail steps encompassing neural network training, an explanation phase, and analysis via feature mapping.For complete details on the use and execution of this protocol, please refer to Mastropietro et al. (2022).1 : Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics. |
first_indexed | 2024-04-11T05:52:56Z |
format | Article |
id | doaj.art-3070fee12fb74adeaf522be6f42ed7cc |
institution | Directory Open Access Journal |
issn | 2666-1667 |
language | English |
last_indexed | 2024-04-11T05:52:56Z |
publishDate | 2022-12-01 |
publisher | Elsevier |
record_format | Article |
series | STAR Protocols |
spelling | doaj.art-3070fee12fb74adeaf522be6f42ed7cc2022-12-22T04:42:00ZengElsevierSTAR Protocols2666-16672022-12-0134101887Protocol to explain graph neural network predictions using an edge-centric Shapley value-based approachAndrea Mastropietro0Giuseppe Pasculli1Jürgen Bajorath2Department of Computer, Control and Management Engineering Antonio Ruberti (DIAG) Sapienza University of Rome, 00185 Rome, Italy; Department of Life Science Informatics and Data Science, B-IT, LIMES Program Unit Chemical Biology and Medicinal Chemistry, Rheinische Friedrich-Wilhelms-Universität, Friedrich-Hirzebruch-Allee 5/6, 53115 Bonn, Germany; Corresponding authorDepartment of Computer, Control and Management Engineering Antonio Ruberti (DIAG) Sapienza University of Rome, 00185 Rome, ItalyDepartment of Life Science Informatics and Data Science, B-IT, LIMES Program Unit Chemical Biology and Medicinal Chemistry, Rheinische Friedrich-Wilhelms-Universität, Friedrich-Hirzebruch-Allee 5/6, 53115 Bonn, Germany; Corresponding authorSummary: Here we present EdgeSHAPer, a workflow for explaining graph neural networks by approximating Shapley values using Monte Carlo sampling. In this protocol, we describe steps to execute Python scripts for a chemical dataset from the original publication; however, this approach is also applicable to any user-provided dataset. We also detail steps encompassing neural network training, an explanation phase, and analysis via feature mapping.For complete details on the use and execution of this protocol, please refer to Mastropietro et al. (2022).1 : Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.http://www.sciencedirect.com/science/article/pii/S2666166722007675BioinformaticsChemistryComputer sciences |
spellingShingle | Andrea Mastropietro Giuseppe Pasculli Jürgen Bajorath Protocol to explain graph neural network predictions using an edge-centric Shapley value-based approach STAR Protocols Bioinformatics Chemistry Computer sciences |
title | Protocol to explain graph neural network predictions using an edge-centric Shapley value-based approach |
title_full | Protocol to explain graph neural network predictions using an edge-centric Shapley value-based approach |
title_fullStr | Protocol to explain graph neural network predictions using an edge-centric Shapley value-based approach |
title_full_unstemmed | Protocol to explain graph neural network predictions using an edge-centric Shapley value-based approach |
title_short | Protocol to explain graph neural network predictions using an edge-centric Shapley value-based approach |
title_sort | protocol to explain graph neural network predictions using an edge centric shapley value based approach |
topic | Bioinformatics Chemistry Computer sciences |
url | http://www.sciencedirect.com/science/article/pii/S2666166722007675 |
work_keys_str_mv | AT andreamastropietro protocoltoexplaingraphneuralnetworkpredictionsusinganedgecentricshapleyvaluebasedapproach AT giuseppepasculli protocoltoexplaingraphneuralnetworkpredictionsusinganedgecentricshapleyvaluebasedapproach AT jurgenbajorath protocoltoexplaingraphneuralnetworkpredictionsusinganedgecentricshapleyvaluebasedapproach |