A key task in structural biology is to define a meaningful similarity measure for the comparison of protein structures. Recently, the use of graphs as modeling tools for molecular data has gained increasing importance. In this context, kernel functions have attracted a lot of attention, especially since they allow for the application of a rich repertoire of methods from the field of kernel-based machine learning. However, most of the existing graph kernels have been designed for unlabeled and/or unweighted graphs, although proteins are often more naturally and more exactly represented in terms of node-labeled and edge-weighted graphs. Here we analyze kernel-based protein comparison methods and propose extensions to existing graph kernels to exploit node-labeled and edge-weighted graphs. Moreover, we propose an instance of the substructure fingerprint kernel suitable for the analysis of protein binding sites. By using fuzzy fingerprints, we solve the problem of discontinuity on bin-boundaries arising in the case of labeled graphs.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1002/minf.201100149 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Transformer Graph Variational Autoencoder for Generative Molecular Design.
Biophys J
January 2025
Department of Machine Learning, Moffitt Cancer Center, Tampa, Florida, United States. Electronic address:
In the field of drug discovery, the generation of new molecules with desirable properties remains a critical challenge. Traditional methods often rely on SMILES (Simplified Molecular Input Line Entry System) representations for molecular input data, which can limit the diversity and novelty of generated molecules. To address this, we present the Transformer Graph Variational Autoencoder (TGVAE), an innovative AI model that employs molecular graphs as input data, thus captures the complex structural relationships within molecules more effectively than string models.
View Article and Find Full Text PDFDevelopment and evaluation of patient-centred polygenic risk score reports for glaucoma screening.
BMC Med Genomics
January 2025
Department of Ophthalmology, Flinders Medical and Health Research Institute, Flinders University, Adelaide, SA, Australia.
Background: Polygenic risk scores (PRS), which provide an individual probabilistic estimate of genetic susceptibility to develop a disease, have shown effective risk stratification for glaucoma onset. However, there is limited best practice evidence for reporting PRS and patient-friendly reports for communicating PRS effectively are lacking. Here we developed patient-centred PRS reports for glaucoma screening based on the literature, and evaluated them with participants using a qualitative research approach.
View Article and Find Full Text PDFA new epidemic model of sexually transmittable diseases: a fractional numerical approach.
Sci Rep
January 2025
Department of Mathematics, College of Science Al-Zulfi, Majmaah University, Al-Majmaah 11952, Saudi Arabia.
This study aims at investigating the dynamics of sexually transmitted infectious disease (STID), which is serious health concern. In so doing, the integer order STID model is progressed in to the time-delayed non-integer order STID model by introducing the Caputo fractional derivatives in place of integer order derivatives and including the delay factors in the susceptible and infectious compartments. Moreover, unique existence of the solution for the underlying model is ensured by establishing some benchmark results.
View Article and Find Full Text PDFExPDrug: Integration of an interpretable neural network and knowledge graph for pathway-based drug repurposing.
Comput Biol Med
January 2025
Department of Computer Engineering, Chungbuk National University, Cheongju, Republic of Korea. Electronic address:
Precision medicine aims to provide personalized therapies by analyzing patient molecular profiles, often focusing on gene expression data. However, effectively linking these data to actionable drug discovery for clinical application remains challenging. In this paper, we introduce ExPDrug, a neural network (NN) model that integrates biological pathways from transcriptomic data with a biomedical knowledge graph to facilitate pathway-based drug repurposing.
View Article and Find Full Text PDFASGCL: Adaptive Sparse Mapping-based graph contrastive learning network for cancer drug response prediction.
PLoS Comput Biol
January 2025
School of Software, Taiyuan University of Technology, Taiyuan, China.
Personalized cancer drug treatment is emerging as a frontier issue in modern medical research. Considering the genomic differences among cancer patients, determining the most effective drug treatment plan is a complex and crucial task. In response to these challenges, this study introduces the Adaptive Sparse Graph Contrastive Learning Network (ASGCL), an innovative approach to unraveling latent interactions in the complex context of cancer cell lines and drugs.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!