CROssBAR: comprehensive resource of biomedical relations with knowledge graph representations

Doğan, Tunca; Ataş, Heval; Joshi, Vishal; Atakan, Ahmet; Rifaioğlu, Ahmet Süreyya; Nalbat, Esra; Nightingale, Andrew; Saidi, Rabie; Volynkin, Vladimir; Zellner, Hermann; Çetin-Atalay, Rengül; Martin, Maria Jesus; Atalay, Volkan

doi:10.1093/nar/gkab543

Cited by 23 publications

(12 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Considering this fact, taking a systems-based approach with the integration and utilization of direct and indirect relationships in molecular and cellular processes including protein–protein interactions, drug/compound–target protein interactions, and signaling/metabolic pathways, together with high level concepts such as protein-disease relationships, drug-disease indications, pathway-disease modulations, and phenotypic implications could increase the success rate in drug discovery. Thus, we aim to construct a new type of systems-level DTI representation and subsequent prediction framework, using CROssBAR [ 54 ] which is an open-source system that integrates large-scale biological/biomedical data and represents them in the form of heterogeneous and computable knowledge graphs. The newly proposed framework will utilize graph representation learning algorithms to process these biomedical knowledge graphs, and will be trained, validated/optimized, and tested on our realistic and challenging datasets.…”

Section: Discussionmentioning

confidence: 99%

How to approach machine learning-based prediction of drug/compound–target interactions

Guvenilir

Doğan

2023

J Cheminform

Self Cite

View full text Add to dashboard Cite

The identification of drug/compound–target interactions (DTIs) constitutes the basis of drug discovery, for which computational predictive approaches have been developed. As a relatively new data-driven paradigm, proteochemometric (PCM) modeling utilizes both protein and compound properties as a pair at the input level and processes them via statistical/machine learning. The representation of input samples (i.e., proteins and their ligands) in the form of quantitative feature vectors is crucial for the extraction of interaction-related properties during the artificial learning and subsequent prediction of DTIs. Lately, the representation learning approach, in which input samples are automatically featurized via training and applying a machine/deep learning model, has been utilized in biomedical sciences. In this study, we performed a comprehensive investigation of different computational approaches/techniques for protein featurization (including both conventional approaches and the novel learned embeddings), data preparation and exploration, machine learning-based modeling, and performance evaluation with the aim of achieving better data representations and more successful learning in DTI prediction. For this, we first constructed realistic and challenging benchmark datasets on small, medium, and large scales to be used as reliable gold standards for specific DTI modeling tasks. We developed and applied a network analysis-based splitting strategy to divide datasets into structurally different training and test folds. Using these datasets together with various featurization methods, we trained and tested DTI prediction models and evaluated their performance from different angles. Our main findings can be summarized under 3 items: (i) random splitting of datasets into train and test folds leads to near-complete data memorization and produce highly over-optimistic results, as a result, should be avoided, (ii) learned protein sequence embeddings work well in DTI prediction and offer high potential, despite interaction-related properties (e.g., structures) of proteins are unused during their self-supervised model training, and (iii) during the learning process, PCM models tend to rely heavily on compound features while partially ignoring protein features, primarily due to the inherent bias in DTI data, indicating the requirement for new and unbiased datasets. We hope this study will aid researchers in designing robust and high-performing data-driven DTI prediction systems that have real-world translational value in drug discovery.

show abstract

Section: Discussionmentioning

confidence: 99%

How to approach machine learning-based prediction of drug/compound–target interactions

Guvenilir

Doğan

2023

J Cheminform

Self Cite

View full text Add to dashboard Cite

show abstract

“…Furthermore, we plan to extend our mappings to uncharacterized protein sequence signatures using sources such as Pfam’s domains of unknown function (DUFs) [ 56 ], and potentially functional regions detected by different computational approaches [ 57 ]. Additionally, we are going to integrate DRUIDom’s compound–domain and compound–target interaction predictions to our large-scale biological and biomedical data integration and representation system CROssBAR [ 58 ] with the aim of enriching the biological relationship-based information provided in this service ( https://crossbar.kansil.org/ ). This way, users can easily browse pre-computed DRUIDom associations/predictions for their proteins of interest, on the fly, together with other types of biomolecular relationships provided in this system (i.e., genes/proteins to diseases, phenotypes, pathways/functions, drugs, in addition to PPIs).…”

Section: Discussionmentioning

confidence: 99%

Protein domain-based prediction of drug/compound–target interactions and experimental validation on LIM kinases

et al. 2021

Self Cite

View full text Add to dashboard Cite

Predictive approaches such as virtual screening have been used in drug discovery with the objective of reducing developmental time and costs. Current machine learning and network-based approaches have issues related to generalization, usability, or model interpretability, especially due to the complexity of target proteins’ structure/function, and bias in system training datasets. Here, we propose a new method “DRUIDom” (DRUg Interacting Domain prediction) to identify bio-interactions between drug candidate compounds and targets by utilizing the domain modularity of proteins, to overcome problems associated with current approaches. DRUIDom is composed of two methodological steps. First, ligands/compounds are statistically mapped to structural domains of their target proteins, with the aim of identifying their interactions. As such, other proteins containing the same mapped domain or domain pair become new candidate targets for the corresponding compounds. Next, a million-scale dataset of small molecule compounds, including those mapped to domains in the previous step, are clustered based on their molecular similarities, and their domain associations are propagated to other compounds within the same clusters. Experimentally verified bioactivity data points, obtained from public databases, are meticulously filtered to construct datasets of active/interacting and inactive/non-interacting drug/compound–target pairs (~2.9M data points), and used as training data for calculating parameters of compound–domain mappings, which led to 27,032 high-confidence associations between 250 domains and 8,165 compounds, and a finalized output of ~5 million new compound–protein interactions. DRUIDom is experimentally validated by syntheses and bioactivity analyses of compounds predicted to target LIM-kinase proteins, which play critical roles in the regulation of cell motility, cell cycle progression, and differentiation through actin filament dynamics. We showed that LIMK-inhibitor-2 and its derivatives significantly block the cancer cell migration through inhibition of LIMK phosphorylation and the downstream protein cofilin. One of the derivative compounds (LIMKi-2d) was identified as a promising candidate due to its action on resistant Mahlavu liver cancer cells. The results demonstrated that DRUIDom can be exploited to identify drug candidate compounds for intended targets and to predict new target proteins based on the defined compound–domain relationships. Datasets, results, and the source code of DRUIDom are fully-available at: https://github.com/cansyl/DRUIDom.

show abstract

“…In further studies, selected de novo molecules will be subjected to chemical synthesis and subsequent in vitro cell-based experiments to validate AKT1 targeting and observe phenotypic effects on HCC celllines. We also plan to improve the molecular generation process by incorporating high-level functional properties of real drugs and drug candidate molecules (along with their structural features, which are already utilized in the current version) in the context of heterogeneous biomedical knowledge graphs [70], to the model training procedure. This architecture is intended to facilitate the understanding of the relationship between the structural and functional properties of small molecules and thereby enhance the design process.…”

Section: Discussionmentioning

confidence: 99%

Target Specific De Novo Design of Drug Candidate Molecules with Graph Transformer-based Generative Adversarial Networks

Ünlü¹,

Çevrim²,

Sarıgün³

et al. 2023

Preprint

View full text Add to dashboard Cite

Discovering novel drug candidate molecules is one of the most fundamental and critical steps in drug development. Generative deep learning models, which create synthetic data given a probability distribution, have been developed with the purpose of picking completely new samples from a partially known space. Generative models offer high potential for designing de novo molecules; however, in order for them to be useful in real-life drug development pipelines, these models should be able to design target-specific molecules, which is the next step in this field. In this study, we propose a novel generative system, DrugGEN, for the de novo design of drug candidate molecules that interact with selected target proteins. The proposed system represents compounds and protein structures as graphs and processes them via serially connected two generative adversarial networks comprising graph transformers. DrugGEN is implemented with five independent models, each with a unique sample generation routine. The system

show abstract

CROssBAR: comprehensive resource of biomedical relations with knowledge graph representations

Cited by 23 publications

References 44 publications

How to approach machine learning-based prediction of drug/compound–target interactions

How to approach machine learning-based prediction of drug/compound–target interactions

Protein domain-based prediction of drug/compound–target interactions and experimental validation on LIM kinases

Target Specific De Novo Design of Drug Candidate Molecules with Graph Transformer-based Generative Adversarial Networks

Contact Info

Product

Resources

About