Classifying breast cancer using multi-view graph neural network based on multi-omics data DOI Creative Commons

Yanjiao Ren,

Yimeng Gao,

Wei Du

et al.

Frontiers in Genetics, Journal Year: 2024, Volume and Issue: 15

Published: Feb. 20, 2024

Introduction: As the evaluation indices, cancer grading and subtyping have diverse clinical, pathological, molecular characteristics with prognostic therapeutic implications. Although researchers begun to study differentiation subtype prediction, most of relevant methods are based on traditional machine learning rely single omics data. It is necessary explore a deep algorithm that integrates multi-omics data achieve classification prediction subtypes. Methods: This paper proposes fusion multi-view graph neural network (MVGNN) for predicting classification. The model framework consists convolutional (GCN) module features from different an attention integrating Three types used. For each type data, feature selection performed using such as chi-square test minimum redundancy maximum relevance (mRMR). Weighted patient similarity networks constructed selected features, GCN trained corresponding networks. Finally, performs final prediction. Results: To validate predictive performance MVGNN model, we conducted experimental comparisons models currently popular 5-fold cross-validation. Additionally, comparative experiments its subtypes two three Discussion: proposed it well in multiple

Language: Английский

Artificial intelligence assists precision medicine in cancer treatment DOI Creative Commons
Jinzhuang Liao, M Kellis, Yu Gan

et al.

Frontiers in Oncology, Journal Year: 2023, Volume and Issue: 12

Published: Jan. 4, 2023

Cancer is a major medical problem worldwide. Due to its high heterogeneity, the use of same drugs or surgical methods in patients with tumor may have different curative effects, leading need for more accurate treatment tumors and personalized treatments patients. The precise essential, which renders obtaining an in-depth understanding changes that undergo urgent, including their genes, proteins cancer cell phenotypes, order develop targeted strategies Artificial intelligence (AI) based on big data can extract hidden patterns, important information, corresponding knowledge behind enormous amount data. For example, ML deep learning subsets AI be used mine deep-level information genomics, transcriptomics, proteomics, radiomics, digital pathological images, other data, make clinicians synthetically comprehensively understand tumors. In addition, find new biomarkers from assist screening, detection, diagnosis, prognosis prediction, so as providing best individual improving clinical outcomes.

Language: Английский

Citations

117
A benchmark study of deep learning-based multi-omics data fusion methods for cancer DOI Creative Commons

Dongjin Leng,

Linyi Zheng,

Yuqi Wen

et al.

Genome biology, Journal Year: 2022, Volume and Issue: 23(1)

Published: Aug. 9, 2022

A fused method using a combination of multi-omics data enables comprehensive study complex biological processes and highlights the interrelationship relevant biomolecules their functions. Driven by high-throughput sequencing technologies, several promising deep learning methods have been proposed for fusing generated from large number samples.

Language: Английский

Citations

78
The Rise of Gastrointestinal Cancers as a Global Phenomenon: Unhealthy Behavior or Progress? DOI Open Access
Sílvia Rodrigues Jardim,

Lucila Marieta Perrotta de Souza,

Heitor Siffert Pereira de Souza

et al.

International Journal of Environmental Research and Public Health, Journal Year: 2023, Volume and Issue: 20(4), P. 3640 - 3640

Published: Feb. 18, 2023

The overall burden of cancer is rapidly increasing worldwide, reflecting not only population growth and aging, but also the prevalence spread risk factors. Gastrointestinal (GI) cancers, including stomach, liver, esophageal, pancreatic, colorectal represent more than a quarter all cancers. While smoking alcohol use are factors most commonly associated with development, growing consensus includes dietary habits as relevant for GI Current evidence suggests that socioeconomic development results in several lifestyle modifications, shifts from local traditional diets to less-healthy Western diets. Moreover, recent data indicate increased production consumption processed foods underlies current pandemics obesity related metabolic disorders, which directly or indirectly emergence various chronic noncommunicable conditions However, environmental changes restricted patterns, unhealthy behavioral features should be analyzed holistic view lifestyle. In this review, we discussed epidemiological aspects, gut dysbiosis, cellular molecular characteristics cancers explored impact behaviors, diet, physical activity on developing context progressive societal changes.

Language: Английский

Citations

71
Multimodal data integration for oncology in the era of deep neural networks: a review DOI Creative Commons
Asim Waqas, Aakash Tripathi, Ravi P. Ramachandran

et al.

Frontiers in Artificial Intelligence, Journal Year: 2024, Volume and Issue: 7

Published: July 25, 2024

Cancer research encompasses data across various scales, modalities, and resolutions, from screening diagnostic imaging to digitized histopathology slides types of molecular clinical records. The integration these diverse for personalized cancer care predictive modeling holds the promise enhancing accuracy reliability screening, diagnosis, treatment. Traditional analytical methods, which often focus on isolated or unimodal information, fall short capturing complex heterogeneous nature data. advent deep neural networks has spurred development sophisticated multimodal fusion techniques capable extracting synthesizing information disparate sources. Among these, Graph Neural Networks (GNNs) Transformers have emerged as powerful tools learning, demonstrating significant success. This review presents foundational principles learning including oncology taxonomy strategies. We delve into recent advancements in GNNs oncology, spotlighting key studies their pivotal findings. discuss unique challenges such heterogeneity complexities, alongside opportunities it a more nuanced comprehensive understanding cancer. Finally, we present some latest pan-cancer By surveying landscape our goal is underline transformative potential Transformers. Through technological methodological innovations presented this review, aim chart course future promising field. may be first that highlights current state applications using transformers, sources, sets stage evolution, encouraging further exploration care.

Language: Английский

Citations

26
A review of cancer data fusion methods based on deep learning DOI
Yuxin Zhao, Xiaobo Li, Changjun Zhou

et al.

Information Fusion, Journal Year: 2024, Volume and Issue: 108, P. 102361 - 102361

Published: March 20, 2024

Language: Английский

Citations

22
Subtype-DCC: decoupled contrastive clustering method for cancer subtype identification based on multi-omics data DOI
Jing Zhao, Bowen Zhao, Xiaotong Song

et al.

Briefings in Bioinformatics, Journal Year: 2023, Volume and Issue: 24(2)

Published: Jan. 25, 2023

Abstract Due to the high heterogeneity and complexity of cancers, patients with different cancer subtypes often have distinct groups genomic clinical characteristics. Therefore, discovery identification are crucial diagnosis, prognosis treatment. Recent technological advances accelerated increasing availability multi-omics data for subtyping. To take advantage complementary information from data, it is necessary develop computational models that can represent integrate layers into a single framework. Here, we propose decoupled contrastive clustering method (Subtype-DCC) based on integration identify subtypes. The idea learning introduced deep neural networks learn clustering-friendly representations. Experimental results demonstrate superior performance proposed Subtype-DCC model in identifying over currently available state-of-the-art methods. strength also supported by survival analysis.

Language: Английский

Citations

26
Challenges and best practices in omics benchmarking DOI
Thomas G. Brooks, Nicholas F. Lahens,

Antonijo Mrčela

et al.

Nature Reviews Genetics, Journal Year: 2024, Volume and Issue: 25(5), P. 326 - 339

Published: Jan. 12, 2024

Language: Английский

Citations

18
A comprehensive review of machine learning techniques for multi-omics data integration: challenges and applications in precision oncology DOI

Debabrata Acharya,

Anirban Mukhopadhyay

Briefings in Functional Genomics, Journal Year: 2024, Volume and Issue: 23(5), P. 549 - 560

Published: April 10, 2024

Multi-omics data play a crucial role in precision medicine, mainly to understand the diverse biological interaction between different omics. Machine learning approaches have been extensively employed this context over years. This review aims comprehensively summarize and categorize these advancements, focusing on integration of multi-omics data, which includes genomics, transcriptomics, proteomics metabolomics, alongside clinical data. We discuss various machine techniques computational methodologies used for integrating distinct omics datasets provide valuable insights into their application. The emphasizes both challenges opportunities present integration, medicine patient stratification, offering practical recommendations method selection scenarios. Recent advances deep network-based are also explored, highlighting potential harmonize information layers. Additionally, we roadmap oncology, outlining advantages, implementation difficulties. Hence offers thorough overview current literature, providing researchers with particularly oncology. Contact: [email protected].

Language: Английский

Citations

18
Molecular similarity: Theory, applications, and perspectives DOI Creative Commons

Kenneth López‐Pérez,

Juan F. Avellaneda-Tamayo, Lexin Chen

et al.

Artificial Intelligence Chemistry, Journal Year: 2024, Volume and Issue: 2(2), P. 100077 - 100077

Published: Aug. 31, 2024

Molecular similarity pervades much of our understanding and rationalization chemistry. This has become particularly evident in the current data-intensive era chemical research, with measures serving as backbone many Machine Learning (ML) supervised unsupervised procedures. Here, we present a discussion on role molecular drug design, space exploration, "art" generation, representations, more. We also discuss more recent topics similarity, like ability to efficiently compare large libraries.

Language: Английский

Citations

12
Artificial intelligence‐driven change redefining radiology through interdisciplinary innovation DOI Creative Commons
Runqiu Huang, Xiaolin Meng, Xiaoxuan Zhang

et al.

Deleted Journal, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 6, 2025

Abstract Artificial intelligence (AI) is rapidly advancing, yet its applications in radiology remain relatively nascent. From a spatiotemporal perspective, this review examines the forces driving AI development and integration with medicine radiology, particular focus on advancements addressing major diseases that significantly threaten human health. Temporally, advent of foundational model architectures, combined underlying drivers development, accelerating progress interventions their practical applications. Spatially, discussion explores potential evolving methodologies to strengthen interdisciplinary within medicine, emphasizing four critical points imaging process, as well application disease management, including emergence commercial products. Additionally, current utilization deep learning reviewed, future through multimodal foundation models Generative Pre‐trained Transformer are anticipated.

Language: Английский

Citations

2