About this Research Topic
In recent years, large amounts of omics (genomics, epigenomics and proteomics, etc.) data, in various forms and with different characteristics, have emerged. This omics data provides us with an unprecedented data resource bearing rich information about biological mechanisms and regulation, however it also presents the challenges of how to mine the information and how to integrate the different forms of data to generate valuable biomedical insights. Machine learning technologies have emerged as a useful tool for improving cancer patient diagnosis and prognosis in clinical settings. Deep learning is a subset of machine learning that has networks capable of learning from large amounts of diverse data with inter-connected relationships. These computational methods can generate insightful hypotheses facilitating experimental validation.
There is still much more to be explored, such as integrating genomics, epigenomics, and proteomics, images, and electronic health records in applications to the studies of cancers. In this Research Topic, we plan to present state-of-the-art machine learning applications for integrative omics in cancer studies, discuss the advantages and limitations of various methods, and, most importantly, the mindset the researchers take to develop a well-performing algorithm. Beyond computational methods, we also want to discuss the experimental validation of these tools for diagnosis and prognosis.
Data of various sources and levels (e.g., gene expression level, methylation level, and protein expression level) may imbed complex relationships. There are many potential challenges in data analysis including limitations to signal detection, missing values, and high-dimensional data structures. Besides computational modeling, experimental validation could also be used to highlight the issues in data analysis in the real world, and help solve them. In the process of experimental validation, we will see whether the computational methods are best suitable for the biological data, and what limitations there are and propose ways to improve the methods.
This Research Topic aims to explore and assess state-of-the-art technologies in machine learning, especially deep learning, for integrating heterogeneous data of different omics, specifically focusing on potential applications for cancer studies. We will focus on both computational methods and wet-lab experimental validation methods, and welcome combination studies.
Areas to be covered in this Research Topic may include, but are not limited to:
• Biomarker discovery
• Cancer heterogeneity
• Drug combination and development
• Image analysis
• Molecular functions
• NLP/electronic health record analysis
This Research Topic welcomes the following article types: Brief Research Report, Editorial, Methods, Mini Review, Opinion, Original Research, Perspective, Review, and Technology and Code.
Keywords: Deep learning, Integrative omics, Cancer, Big data, Artificial intelligence
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.