FP-GNN: a versatile deep learning architecture for enhanced molecular property prediction

Cai, Hanxuan; Zhang, Huimin; Zhao, Duancheng; Wu, Jianhui

doi:10.48550/arxiv.2205.03834

Cited by 2 publications

(7 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…LIT-PCBA [37] is a virtual screening dataset containing 15 protein targets, 9780 active compounds (positive samples), and 407,839 unique inactive compounds (negative samples) selected from high-confidence PubChem Bioassay data. Predicting GEM-2 A PREPRINT a These results are collected from [5], where standard deviations are not reported.…”

Section: Drug Discovery Benchmarkmentioning

confidence: 99%

“…Due to the large prediction variance on the classification tasks with only a few positive samples, we only evaluate the methods on the targets (ALDH1, FEN1, GBA, KAT2A, MAPK1, PKM2, and VDR) with more than 150 active compounds. Following the previous work [5], we split the samples of each protein target into the training and test sets at the ratio of 3:1 with asymmetric validation embedding (AVE) method [39]. The split dataset can be directly downloaded at the previous work's GitHub repository 5 .…”

Section: Drug Discovery Benchmarkmentioning

confidence: 99%

“…Baselines. Similar to the experiments on PCQM4Mv2, we compare GEM-2 with multiple types of baselines: (1) traditional machine learning methods, including Naive Bayes(NB) [11], support vector machine (SVM) [8], random forest (RF) [27] and extreme gradient boosting (XGBoost) [7]; (2) w/o full-range & w/o many-body methods, including GCN, Graph Attention Network (GAT) [38], and FP-GNN [5]; (3) w/ full-range & w/o many-body methods, including EGT and EGT pretrain [18]; (4) w/o full-range & w/ many-body methods, including GEM and GEM pretrain [13]. GEM-2 and GEM-2 pretrain are our proposed methods.…”

Section: Drug Discovery Benchmarkmentioning

confidence: 99%

“…To compare the accuracy of GEM-2 and the baseline methods, we use ROC-AUC (area under the receiver operating characteristic curve) as the evaluation metric. The results of NB, SVM, RF, XGBoost, GCN, GAT, and FP-GNN are collected from [5]. The results of EGT, EGT pretrain , GEM, GEM pretrain are obtained by running the released code with hyper-parameter searching on dataset LIT-PCBA.…”

Section: Drug Discovery Benchmarkmentioning

confidence: 99%

“…For Quantum Chemistry Benchmark, PCQM4Mv2 dataset is publicly available on OGB's official website(https:// ogb.stanford.edu/docs/lsc/pcqm4mv2/). For Drug Discovery Benchmark, LIT-PCBA dataset can be downloaded from [5]'s Github repository(https://github.com/idrugLab/FP-GNN/blob/main/Data.rar).…”

Section: Data Availabilitymentioning

confidence: 99%

See 4 more Smart Citations

GEM-2: Next Generation Molecular Property Prediction Network by Modeling Full-range Many-body Interactions

Liu

Fang

et al. 2022

Preprint

View full text Add to dashboard Cite

Molecular property prediction is a fundamental task in the drug and material industries. Physically, the properties of a molecule are determined by its own electronic structure, which is a quantum many-body system and can be exactly described by the Schrödinger equation. Full-range many-body interactions between electrons have been proven effective in obtaining an accurate solution of the Schrödinger equation by classical computational chemistry methods, although modeling such interactions consumes an expensive computational cost. Meanwhile, deep learning methods have also demonstrated their competence in molecular property prediction tasks. Inspired by the classical computational chemistry methods, we design a novel method, namely GEM-2, which comprehensively considers full-range many-body interactions in molecules. Multiple tracks are utilized to model the full-range interactions between the many-bodies with different orders, and a novel axial attention mechanism is designed to approximate the full-range interaction modeling with much lower computational cost.

show abstract

Section: Drug Discovery Benchmarkmentioning

confidence: 99%

Section: Drug Discovery Benchmarkmentioning

confidence: 99%

Section: Drug Discovery Benchmarkmentioning

confidence: 99%

Section: Drug Discovery Benchmarkmentioning

confidence: 99%

Section: Data Availabilitymentioning

confidence: 99%

See 3 more Smart Citations

GEM-2: Next Generation Molecular Property Prediction Network by Modeling Full-range Many-body Interactions

Liu

Fang

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

GraphEGFR: Multi‐task and transfer learning based on molecular graph attention mechanism and fingerprints improving inhibitor bioactivity prediction for EGFR family proteins on data scarcity

Boonyarit,

Yamprasert,

Kaewnuratchadasorn

et al. 2024

J Comput Chem

View full text Add to dashboard Cite

The proteins within the human epidermal growth factor receptor (EGFR) family, members of the tyrosine kinase receptor family, play a pivotal role in the molecular mechanisms driving the development of various tumors. Tyrosine kinase inhibitors, key compounds in targeted therapy, encounter challenges in cancer treatment due to emerging drug resistance mutations. Consequently, machine learning has undergone significant evolution to address the challenges of cancer drug discovery related to EGFR family proteins. However, the application of deep learning in this area is hindered by inherent difficulties associated with small‐scale data, particularly the risk of overfitting. Moreover, the design of a model architecture that facilitates learning through multi‐task and transfer learning, coupled with appropriate molecular representation, poses substantial challenges. In this study, we introduce GraphEGFR, a deep learning regression model designed to enhance molecular representation and model architecture for predicting the bioactivity of inhibitors against both wild‐type and mutant EGFR family proteins. GraphEGFR integrates a graph attention mechanism for molecular graphs with deep and convolutional neural networks for molecular fingerprints. We observed that GraphEGFR models employing multi‐task and transfer learning strategies generally achieve predictive performance comparable to existing competitive methods. The integration of molecular graphs and fingerprints adeptly captures relationships between atoms and enables both global and local pattern recognition. We further validated potential multi‐targeted inhibitors for wild‐type and mutant HER1 kinases, exploring key amino acid residues through molecular dynamics simulations to understand molecular interactions. This predictive model offers a robust strategy that could significantly contribute to overcoming the challenges of developing deep learning models for drug discovery with limited data and exploring new frontiers in multi‐targeted kinase drug discovery for EGFR family proteins.

show abstract

FP-GNN: a versatile deep learning architecture for enhanced molecular property prediction

Cited by 2 publications

References 0 publications

GEM-2: Next Generation Molecular Property Prediction Network by Modeling Full-range Many-body Interactions

GEM-2: Next Generation Molecular Property Prediction Network by Modeling Full-range Many-body Interactions

GraphEGFR: Multi‐task and transfer learning based on molecular graph attention mechanism and fingerprints improving inhibitor bioactivity prediction for EGFR family proteins on data scarcity

Contact Info

Product

Resources

About