Due to the fast-developing events surrounding COVID-19 and growing concern for maintaining our collective health, we have decided to cancel the Nexus Informatics Conference scheduled at Kansas City University for April 16 and 17, 2020.
THURSDAY, APRIL 16, 2020
10:00 AM–10:30 AM
WELCOME AND INTRODUCTIONS
Dennis Ridenour | Welcome
10:30 AM – 11:30
Subha Madhavan, PhD, FACMI | Virtual Molecular Tumor Boards – A Global Sharing of Genomic Expertise in Clinical Practice
The cancer research community is on the precipice of transformational advancements in understanding tumor biology, disease etiology, risk stratification, and pathways to novel treatments. Yet the clinical cancer genomics field has been stymied by siloed efforts to meaningfully collect, interpret and aggregate disparate data types from multiple high-throughput modalities into clinical care processes. Many informatics efforts focused on genomic interpretation resources for neoplasms are underway to support case review and treatment planning. Molecular Tumor Boards (MTBs) are a collaborative effort of multi-disciplinary cancer experts equipped with sophisticated genomic interpretation tools to aid in the delivery of accurate and timely clinical interpretations of complex genomic results for each case, within an institution or hospital network. Virtual Molecular Tumor Boards (VMTBs) provide an online forum for collaborative governance, provenance and information sharing between experts around the world. Knowledge sharing in VMTBs and constant communication with guideline developing organizations can lead to progress evidenced by data harmonization across multiple resources, crowd-sourced and expert-curated genomic assertions, and a more informed and explainable usage of artificial intelligence in such settings. Advances in cancer genomics interpretation aid in better patient and disease classification, more streamlined identification of relevant literature, and a more thorough review of available treatments and predicted patient outcomes.
11:30 AM–1:00 PM
LUNCH & POSTER PRESENTATION
1:00 PM – 2:00 PM
MODERATOR: Robert White, PhD, Kansas City University
Ehab Sarsour, PhD | Application of Bioinformatics in Translational Cancer Research
Utilizing multiple sources of bioinformatics can be a very powerful tool for advancing translational cancer research. This presentation will showcase the use of bioinformatics generated from Multi-Omics (genomics, proteomics and metabolomics) combined with bioinformatics from patient’s data banks to decipher mechanisms contributing to therapy resistance of pancreatic and head and neck cancers. Phase I clinical trial showed that head and neck cancer subjects with higher proliferative cancer had a better therapy outcome compared to subjects with a lower proliferative cancer. Bed to bench approach with the integration of bioinformatics was used to investigate these intriguing results and find a solution for this clinical problem of resistance to therapy in patients with less proliferative cancer. Bioinformatics from multi-omics and patient data was also applied to investigate the effects of aging stromal microenvironment on pancreatic cancer progression and response to therapy. Bioinformatics using multi-omics approach was the solution to the challenge in understanding the effects of aging normal cells on the cancer cell microenvironment, which currently lacks patient’s data; most samples only contain cancer cells that do not fully represent the cancer microenvironment. Two different cancers and two different approaches, but in both bioinformatics were the tools that were used to bridge basic science and clinical results that can be considered for translational application.
Abu Mosa, PhD
2:00 PM – 3:00 PM
DATA STANDARDIZATION AND INTEGRATION
MODERATOR: Susan Brown, PhD, Kansas State University
Keith Feldman, PhD | The Healthcare Data Spectrum – Utilizing Multiple Data Sources to Address Novel Healthcare Questions
Coupled with the rise of data science and machine learning, the increasing availability of digitized health and wellness data has provided an exciting opportunity for complex analyses of problems throughout the healthcare domain. Whereas many early works focused on a particular aspect of care, often drawing on data from a specific clinical or administrative data source, it has become clear such a narrow approach is insufficient to capture the complexity of the human condition. Instead, adequately modeling health and wellness problems requires the ability to draw upon data spanning multiple facets of an individual’s biology, their care, and the broader world in which they live.
Accordingly, this talk will present the Healthcare Data Spectrum – a novel conceptual framework designed to provide a comprehensive perspective of health and wellness data as it pertains to an individual. In doing so, it will highlight a few benefits, and challenges, associated with integrating various data sources with regards to both clinical and population health analyses.
Dario Gehrsi, MD, PhD | Computational Methods for Extracting and Visualizing Information from Complex Biological Datasets
In this talk I will discuss analytical tools for working with controlled vocabularies in biology. In particular, I will introduce FunSet, a recently developed command-line tool and web server to perform Gene Ontology enrichment analysis. FunSet allows biologists to: (1) identify terms that are statistically overrepresented in a gene list with respect to a background set; (2) cluster enriched terms using a semantic similarity measure; (3) identify an optimal number of clusters; and (4) display the results using a 2D network view. The talk will also briefly highlight other uses of controlled vocabularies and will touch on some of the limitations of existing approaches.
Dinesh Pal Mudaranthakam, MS, MBA | Automated Data transfer from the Electronic Health Record to Electronic Data Capture System
Research plays a vital role in discovering groundbreaking medications and treatments. Time is of the essence when it comes to oncology since oncology patients need medication and treatment now! One can’t wait for eight or ten years. One of the major elements of conducting a clinical trial is capturing data in Electronic Case Report Forms (eCRFs), which is then analyzed and used as a proof to justify the need for new drug or treatment. KUMC has taken a new approach to autopopulate eCRFs by utilizing a Cancer Curated Clinical Outcomes database (C3OD). Literature suggests countries such as Belgium, London, France, have already tried implementing this approach. Our contribution to the literature is not just auto-populating the forms for our internal needs but also to automatically push and disseminate data to the sponsor (Industry sponsor). Results from our approach suggest a two-prong approach where the structured data elements are straight forward to transmit and populate. The free text (or unstructured) data elements would need some manual verification or a Natural Language Processing (NLP) methodology to interpret/parse the data to answer study-specific questions. Depending upon the phase, design and disease of the clinical trial, we noticed a significant improvement in the quality of data and cost/time savings.
Sonny Lee, PhD | The Impact of Microbiota in Early Life on IBD Development: From Metabarcoding to Metagenomes
Epidemiological studies identified associations between antibiotics exposure in early life and increased risk of developing inflammatory bowel diseases (IBD) later in life. We demonstrated a causative relationship between peripartum antibiotic-induced gut dysbiosis and IBD onset in a murine model. We further examined gut microbiota community membership and functional potential over time using metagenomic shotgun. We found that maternal peripartum antibiotic exposure restructures both the bacteriome and mycobiome membership and alters the functions of the whole microbiome community in offspring relative to those observed in unexposed pups. Among offspring from antibiotic-exposed dams, differences in gene profiles related to nitrogen metabolism were observed between pups that eventually developed overt spontaneous colitis versus those that did not. In the course of disease, community-wide differences in nitrogen metabolism appeared at 8 weeks after withdrawal of antibiotic treatment (11 weeks of age), which was prior to the onset of clinical symptoms. We then performed ex-vivo experiments examining the metabolizing capacity associated with the computationally predicted pathway using living microbiota samples to obtain physiological insights into these differences beyond genetic information. Our study demonstrated that the functional changes of the gut microbiota related to nitrogen metabolism precedes the clinical onset of colitis.
3:00 PM – 3:30 PM
3:30 PM – 4:30 PM
MODERATOR: Martina Clarke, PhD, University of Nebraska Medical Center
Bartosz Grobelny, MD | Imaging the Electrical Connections of the Brain in Surgical Epilepsy Patients as it Relates to Seizure Outcome
Objective: We sought to determine whether the presence or surgical removal of certain nodes in a connectivity network constructed from intracranial electroencephalography recordings determines postoperative seizure freedom in surgical epilepsy patients.
Methods: We analyzed connectivity networks constructed from peri-ictal intracranial electroencephalography of surgical epilepsy patients before a tailored resection. Thirty-six patients and 123 seizures were analyzed. Their Engel class postsurgical seizure outcome was determined at least one year after surgery. Betweenness centrality, a measure of a node’s importance as a hub in the network, was used to compare nodes.
Results: The presence of larger quantities of high-betweenness nodes in interictal and postictal networks was associated with failure to achieve seizure freedom from the surgery(p<0.001), as was resection of high-betweenness nodes in three successive frequency groups in mid-seizure networks(p<0.001).
Conclusions: Betweenness centrality is a biomarker for postsurgical seizure outcomes. The presence of high-betweeeness nodes in interictal and postictal networks can predict patient outcome independent of resection. Additionally, since their resection is associated with worse seizure outcomes, the mid-seizure network high-betweenness centrality nodes may represent hubs in self-regulatory networks that inhibit or help terminate seizures.
Sonya Bahar, PhD | Visualizing Neural Synchronization in the Neocortex
Seizures have typically been understood to involve a dramatic increase in synchronization of neural activity. In order to quantify this synchronization change, we have used voltage-sensitive dye imaging to record the spatial distribution of activity during focal seizures in the rat neocortex, induced by the injection of 4-aminopyridine. I will discuss the pathway from the raw imaging data, recorded using a CCD camera, to visualization and quantification of synchronization changes that correlate with the onset and spread of the epileptiform activity.
4:30 PM – 5:30 PM
MODERATOR: Ryan Moog, Cerner
Bethany Lowndes | Work System Design for Occupational Health and Safety
In many professions, the workplace can be described as a complex sociotechnical system. Human factors theories, principles, and models are applied to examine the relationship between people and these complex work systems. User-centered approaches in the design of workspaces, instrumentation, technology, and organizational practices optimize performance and minimize compromise to health and safety. By understanding human capability, demands of the task, and risks associated with the work, human factors engineers make adaptations to the system in order to improve user—and overall system—performance. Human factors analysis of work systems will be described for occupations in healthcare and agriculture. Methods for data collection and analysis will prioritize occupational health and safety.
Shellie Ellis | Precision Community: Improving Targeted Cancer Therapy Use in Community Oncology
Lindsay Jarrett, PhD | What We Ought to Do With AI in Healthcare
AI systems are becoming increasingly prevalent in the world of healthcare. These complex tools have great potential to improve health outcomes, organizational efficiency, and patient experience, but they also present certain risks. AI systems have been shown to contribute to the reproduction of systemic bias and discrimination in non-healthcare activities like hiring and sentencing, and it is not unreasonable to assume the same could happen in the healthcare setting, a place where constituents are especially vulnerable and where trust is critical. Addressing these risks will require an understanding of the ethical dimensions of AI in healthcare and a community-wide effort to develop and implement AI systems in healthcare in an ethically responsible way. To this end, in 2019 Cerner partnered with the Center for Practical Bioethics to conduct a community-based initiative exploring these ethical dimensions and improving adherence to a broadly accepted ethical framework. This talk will address preliminary findings from a recent KC based Ethical AI workshop, along with recommendations for next steps and future initiatives.
5:30 PM – 8:00PM
COCKTAILS, NETWORKING, & DINNER
FRIDAY, APRIL 17, 2020
7:15 AM–7:45 AM
7:45 AM–8:00 AM
WELCOME & INTRODUCTIONS
Keith Gary, PhD | Thank Sponsors and Volunteers
Mark Hoffman, PhD | Welcome
8:00 AM – 9:00 AM
Casey Overby Taylor, PhD | Precision Medicine Applications with Real World Data
Data from the healthcare delivery system as well as from other sources such as genomics research and digital devices are important drivers of precision medicine. The prospect of computational, model-driven guidance based on genomics and other personal data is growing as the clinical and research enterprise becomes more connected. The ability to implement innovative approaches to guide precision medicine delivery, however, relies on our ability to respond to a changing landscape of patient populations, data collection, data access, model-driven insights, etc. Open source Web services for clinical decision support (CDS) are a promising mechanism to deliver recommendations in a way that can change as evidence derived from multiple data sources matures. This talk will explore ways that data science, biomedical informatics, and implementation science research are helping to realize the potential to use open source Web services for CDS to deliver insights from real world data in a way that is clinically actionable.
9:00 AM – 10:00 AM
MODERATOR: Sharlee Climer, PhD, University of Missouri St. Louis
Monica Gaddis, PhD | The Electronic Health Record: Building Bridges not Walls between the Informaticist and the Clinician
Brad Olson, PhD | Predictive Modeling of Multicellular Evolution
The evolution of multicellular organisms is a major transition of biological complexity that has occurred multiple times on Earth. However, the theoretical and molecular basis for why multicellular organisms have evolved multiple times on this planet is poorly understood. Using large scale “-omics” data sets from the volvocine algae, who recently evolved multicellularity, we are developing predictive models for genes and pathways that are important for evolving multicellular complexity. Model development requires an end-to-end solution that presents multiple challenges to model development. Unsupervised feature engineering of a high dimensional data is the biggest challenge to model development. Current model development is focused on a combining k-nearest neighbor clusters combined with multilayered convolutional neural networks. These layers are then fed into Bayesian neural network for predictions. Models are then trained with “-omics” data from wild type unicellular and multicellular organisms as well as similar data obtained from unicellular mutants of undifferentiated multicellular organisms. This talk will focus on overcoming the challenges of model development with highly dimension data with low replication.
10:00 AM – 10:30 AM
10:30 AM–11:30 AM
MODERATOR: Noah Fahlgren, PhD, Donald Danforth Plant Science Center
Mao Li | Comprehensive 3D Imaging Analysis and Modeling for Inflorescence Architecture
Inflorescences are among the most intensely-studied structures in plant biology due to their importance in not only ecology and evolution, but also agriculture and crop production. However, they contain substantial morphological diversity and complexity, bringing challenges of comprehensive and precise quantification. Most features of inflorescence are not well-measured manually or with standard 2D imaging. Therefore, the combination of new technologies and computational tools is crucial for 3D in-depth phenotyping. In this talk, I will give an example of advanced 3D phenomics. We used X-ray computed tomography, developed custom computational pipelines, and advanced morphometrics to accurately and comprehensively phenotype grapevine inflorescence architecture with both general and organ-specific traits in a 3D framework. This 3D phenomics enable comprehensive characterization of complex architecture and can be broadly applied to root and shoot structures of many species.
Kangwon Seo, PhD | Tracking Alzheimer’s Disease Progression Path in a Two-Dimensional Map
In this presentation, we introduce an assistant tool for Alzheimer’s disease (AD) diagnosis and prognosis by summarizing multiple complicated features from tomographic neuroimages. This tool provides high diagnostic accuracy and sensitivity in tracking Alzheimer’s disease (AD) progression over time in clinical setting. It is also practically intuitive and explainable to patients and their families. This new visualization tool is derived from the manifold-based nonlinear dimension reduction of brain MRI features. In specific, we investigate the locally linear embedding (LLE) method using a dataset from Alzheimer’s Disease Neuroimaging Initiative (ADNI), which includes the longitudinal MRIs from 562 subjects. About 20% of them progressed to the next stage of dementia. Using only the baseline data of cognitively unimpaired and AD subjects, LLE reduces the feature dimension to two and a subject’s AD progression path can be plotted in this low dimensional LLE feature space. In addition, the likelihood of being categorized to AD is indicated by color. This LLE map is a new data visualization tool that can assist in tracking AD progression over time.
11:30 AM–12:30 PM
MODERATOR: Donna Buchanan, PhD, Saint Luke’s
Midhat Farooqi, MD, PHD | A Novel Application of Cerner PowerForms as a Means to Reduce Ordering Errors in Complex Genetic Testing
Commonly used electronic medical record (EMR) platforms demonstrate a lack of conditionality, which becomes increasingly burdensome as laboratory testing becomes more complex. The Center for Pediatric Genomic Medicine (CPGM) at Children’s Mercy Hospital began offering tumor+normal whole genome sequencing for children with cancer in February 2020. This testing involves sequencing the genome of both neoplastic cells and the patient’s germline and then comparing the two to detect somatic variants. Such testing requires collection of multiple sample types from the individual, which vary depending upon the patient’s diagnosis, transplant status, and disease time point. In some cases, DNA from a bone marrow donor or archived tumor specimen must be procured as part of the genetic testing process. Errors in this process were expected to be frequent based on past experience and the complex nature of this testing, and would lead to the recollection of specimens, delays in testing, and decreased patient/provider satisfaction with the genetic testing process. Here, we present a novel application of Cerner PowerForms as a means of obtaining the proper specimens, consent, and preauthorization needed for complex genetic testing without undue burden or educational requirements for ordering providers.