On QSAR-based cardiotoxicity modeling with the expressiveness-enhanced graph learning model and dual-threshold scheme
Copyright © 2023 Wang, Zhu, Izu, Chen-Izu, Ono, Altaf-Ul-Amin, Kanaya and Huang..
Introduction: Given the direct association with malignant ventricular arrhythmias, cardiotoxicity is a major concern in drug design. In the past decades, computational models based on the quantitative structure-activity relationship have been proposed to screen out cardiotoxic compounds and have shown promising results. The combination of molecular fingerprint and the machine learning model shows stable performance for a wide spectrum of problems; however, not long after the advent of the graph neural network (GNN) deep learning model and its variant (e.g., graph transformer), it has become the principal way of quantitative structure-activity relationship-based modeling for its high flexibility in feature extraction and decision rule generation. Despite all these progresses, the expressiveness (the ability of a program to identify non-isomorphic graph structures) of the GNN model is bounded by the WL isomorphism test, and a suitable thresholding scheme that relates directly to the sensitivity and credibility of a model is still an open question. Methods: In this research, we further improved the expressiveness of the GNN model by introducing the substructure-aware bias by the graph subgraph transformer network model. Moreover, to propose the most appropriate thresholding scheme, a comprehensive comparison of the thresholding schemes was conducted. Results: Based on these improvements, the best model attains performance with 90.4% precision, 90.4% recall, and 90.5% F1-score with a dual-threshold scheme (active: <1μM; non-active: >30μM). The improved pipeline (graph subgraph transformer network model and thresholding scheme) also shows its advantages in terms of the activity cliff problem and model interpretability.
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2023 |
|---|---|
| Erschienen: |
2023 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:14 |
|---|---|
| Enthalten in: |
Frontiers in physiology - 14(2023) vom: 19., Seite 1156286 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Wang, Huijia [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
Cardiotoxicity |
|---|
| Anmerkungen: |
Date Revised 06.12.2023 published: Electronic-eCollection Citation Status PubMed-not-MEDLINE |
|---|
| doi: |
10.3389/fphys.2023.1156286 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM357311809 |
|---|
| LEADER | 01000naa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM357311809 | ||
| 003 | DE-627 | ||
| 005 | 20231226072317.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 231226s2023 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.3389/fphys.2023.1156286 |2 doi | |
| 028 | 5 | 2 | |a pubmed24n1190.xml |
| 035 | |a (DE-627)NLM357311809 | ||
| 035 | |a (NLM)37228825 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Wang, Huijia |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a On QSAR-based cardiotoxicity modeling with the expressiveness-enhanced graph learning model and dual-threshold scheme |
| 264 | 1 | |c 2023 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Revised 06.12.2023 | ||
| 500 | |a published: Electronic-eCollection | ||
| 500 | |a Citation Status PubMed-not-MEDLINE | ||
| 520 | |a Copyright © 2023 Wang, Zhu, Izu, Chen-Izu, Ono, Altaf-Ul-Amin, Kanaya and Huang. | ||
| 520 | |a Introduction: Given the direct association with malignant ventricular arrhythmias, cardiotoxicity is a major concern in drug design. In the past decades, computational models based on the quantitative structure-activity relationship have been proposed to screen out cardiotoxic compounds and have shown promising results. The combination of molecular fingerprint and the machine learning model shows stable performance for a wide spectrum of problems; however, not long after the advent of the graph neural network (GNN) deep learning model and its variant (e.g., graph transformer), it has become the principal way of quantitative structure-activity relationship-based modeling for its high flexibility in feature extraction and decision rule generation. Despite all these progresses, the expressiveness (the ability of a program to identify non-isomorphic graph structures) of the GNN model is bounded by the WL isomorphism test, and a suitable thresholding scheme that relates directly to the sensitivity and credibility of a model is still an open question. Methods: In this research, we further improved the expressiveness of the GNN model by introducing the substructure-aware bias by the graph subgraph transformer network model. Moreover, to propose the most appropriate thresholding scheme, a comprehensive comparison of the thresholding schemes was conducted. Results: Based on these improvements, the best model attains performance with 90.4% precision, 90.4% recall, and 90.5% F1-score with a dual-threshold scheme (active: <1μM; non-active: >30μM). The improved pipeline (graph subgraph transformer network model and thresholding scheme) also shows its advantages in terms of the activity cliff problem and model interpretability | ||
| 650 | 4 | |a Journal Article | |
| 650 | 4 | |a cardiotoxicity | |
| 650 | 4 | |a dual-threshold | |
| 650 | 4 | |a graph transformer neural network | |
| 650 | 4 | |a hERG | |
| 650 | 4 | |a meta-path | |
| 700 | 1 | |a Zhu, Guangxian |e verfasserin |4 aut | |
| 700 | 1 | |a Izu, Leighton T |e verfasserin |4 aut | |
| 700 | 1 | |a Chen-Izu, Ye |e verfasserin |4 aut | |
| 700 | 1 | |a Ono, Naoaki |e verfasserin |4 aut | |
| 700 | 1 | |a Altaf-Ul-Amin, M D |e verfasserin |4 aut | |
| 700 | 1 | |a Kanaya, Shigehiko |e verfasserin |4 aut | |
| 700 | 1 | |a Huang, Ming |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Frontiers in physiology |d 2010 |g 14(2023) vom: 19., Seite 1156286 |w (DE-627)NLM205532799 |x 1664-042X |7 nnns |
| 773 | 1 | 8 | |g volume:14 |g year:2023 |g day:19 |g pages:1156286 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.3389/fphys.2023.1156286 |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 14 |j 2023 |b 19 |h 1156286 | ||