Call for papers: Deep phenotyping for Precision Medicine
• 大类 : 医学 - 3区
• 小类 : 计算机：跨学科应用 - 3区
• 小类 : 医学：信息 - 3区
We invite submissions for a special issue of the Journal of Biomedical Informatics focused on Deep Phenotyping to enable Precision Medicine. This special issue aims to provide a collection of emerging theories, cutting-edge methodologies, and novel technologies that enable scal- able human phenotype discovery and application in clinical data for continuous health learning.
An important goal of Precision Medicine is to develop a framework for creating a new taxonomy of human diseases based on molecular biology and then to create it . Towards this goal, clinical data have been recognized as the basic staple of health learning . The rapidly growing interoperable clinical datasets, including electronic health re- cords (EHR), patient-generated self-tracking data, administrative and claims records, and clinical research results data, have presented un- precedented opportunities for developing high-throughput methods for deep phenotyping.
In the context of this special issue, phenotype is “the trait or ob- servable characteristic of a human being representing his or her unique morphological, biochemical, physiological, or behavioral property” . Related, phenotyping refers to the process of characterization or clas- sification of a patient’s phenotype. Deep phenotyping further empha- sizes the precision and comprehensiveness of the characterized phe- notype . Fundamental to studying disease similarities to assist in the development of a precise disease taxonomy, deep phenotyping can shed light on gene functions and enable precise diagnoses, subtyping, and treatments. Software or algorithms leveraging deep phenotyping for gene prioritization have evolved to the point of demonstrating their usefulness in genomic diagnostic decision support [5–13].
Possible topics include, but are not limited to:
Computational phenotype analysis (e.g., causal  or probabilistic phenotyping )
Deep learning for phenotyping
Next-generation phenotyping of electronic health records 
Population Physiology using electronic health records 
High-fidelity phenotyping 
Natural language processing for phenotyping using textual content
Using novel data sources (e.g., patient self-reported data, case re- ports, social media, or clinical research data) for deep phenotyping
Multi-modality phenotyping combining data, text, videos, images, sound, etc.
Integrating multiple data or knowledge sources for phenotype knowledge engineering
Cross-species phenotype knowledge discovery and engineering
Human Phenotype Ontology enhancement or applications
Phenotype-driven disease taxonomy development
Phenotype-driven disease diagnoses or subtyping
Issues and methods for improving the portability of phenotyping methods
Standards-based representation, sharing, and reuse of phenotyping algorithms
Methods for engaging domain experts in high-throughput deep phenotyping