THURSDAY, APRIL 13, 2017 | UMKC STUDENT UNION AUDITORIUM
10:00 AM–10:30 AM
WELCOME AND INTRODUCTIONS
Lawrence Dreyfus, PhD | Welcome
Mark Hoffman, PhD | Working in the Big Tent of Bioinformatics
Keith Gary, PhD | Guidebook App
10:30 AM – 11:30 AM
Liz Worthey, PhD | Making a Definitive Diagnosis; Clinical Application of Whole Genome Sequencing
Despite significant advances in our understanding of the basis of human disease, until recently we have been unable to diagnose a large percentage of patients. Identification of molecular changes provides an opportunity to understand their role in health and disease, and in a clinical setting to apply that understanding to prevention, diagnosis, and treatment. The introduction of clinical genome sequencing has fundamentally altered how molecular changes are identified and has transformed the practice of medicine. It has uncovered disease associated molecular changes that would not otherwise have been identified and has more than tripled the diagnostic success rate in patients with rare genetic diseases. It also reduces the cost of obtaining a definitive diagnosis in rare genetic disease, which in turn reduces the patient care costs. This talk will focus on all of these aspects relating to how we are using whole genome sequencing in a clinical setting.
11:30 AM–1:00 PM
LUNCH & POSTER PRESENTATION | Room 401
1:00 PM – 2:00 PM
MODERATOR: Susan Brown, PhD
Gerald Wyckoff, PhD | Integrating Data in a One Health Project- Challenges and Opportunities
There is currently no accepted structure or best practices for accommodating data originating across human and animal health settings. This is leading to a “gap” in our ability to share data among researchers who are working on essentially the same problem across species- not just between humans and animals, but among disparate animal species. We look at one set of problems originating from the congruence of genetic data across Mendelian diseases as an example of these problems and a model for how to solve them, and urge the creation of a set of best practices that can be followed moving forward. This dovetails with the implementation of the Structured Environment for Animal Data and Simulation (SEADS) project between UMKC and KSU-Olathe.
Zhuqi Miao, PhD| Data Cleaning Problems and Algorithms for Cerner Health Facts Data
Rich health data created via electronic health record (EHR) systems is stimulating the rapid development of EHR-based health research to enhance health-care delivery and practice. Data cleaning is a vital step for the application of data mining in EHR-based health research. Using EHR data extracted from Cerner Health Facts, one of the nation’s largest EHR data warehouses, we identiﬁed prominent data cleaning challenges and developed algorithms to address these issues. Given the considerable similarities among various EHR systems, it is expected that the ﬁndings and algorithms based on Health Facts can be extended to cleaning EHR data in general.
Toni Kazic | Modelling Complex Phenotypes: Tales from the Edges of High-Dimensional Space
Complex phenotypes are very important and very difficult to dissect. Development, diseases, and crop yields in changing climates exemplify traits that are produced by many genes interacting nonlinearly with each other, their products, and their environments to produce a wide and often surprising range of phenotypes. Optimizing therapeutics and crops requires we understand the underlying network and its nonlinear responses. Today, as high throughput phenomics pushes biomedical research into big data, our ability to collect data out-runs our ability to extract the biological causation from this information. Good models capture our current understanding and expose its weaknesses, but are challenging to develop and test for complex phenotypes. I discuss how we have modelled a complex phenotype in maize, including confusions and pratfalls.
Student Presentation: Richard Meier| Digit: A Tool for (D)etection and (I)dentification of (G)enomic (I)nterchromosomal (T)ranslocations
Structural variations (SVs) in genomic DNA can have profound effects on the evolution of living organisms, on phenotypic variations, and on disease processes. A critical step in discovering the full extent of structural variations is the development of tools to characterize these variations accurately in next generation sequencing data. Mate-pair sequencing allows to achieve this goal by identifying clusters of discordantly mapping reads, while also requiring lower read depths than other paired-end strategies. Unfortunately, many of the algorithms designed for PE data fail to address the high number of artifacts generated by mate-pair sequencing due to library preparation. We developed a software pipeline named digit that implements a novel measure of mapping ambiguity to discover interchromosomal SVs from mate-pair and pair-end sequencing data. The workflow robustly handles the high numbers of artifacts present in mate-pair sequencing and reduces the false positive rate while maintaining sensitivity. The workflow recovered 96% of simulated SVs, as well as some selected, previously validated translocations from real datasets. It also generates a self-updating library of common translocations and allows for the investigation of patient- or group-specific events. Utilizing this strategy, a set of translocations potentially associated with lung cancer was separated from common, unspecific events. Identification of group specific events enables discovering and cataloging chromosomal translocations associated with specific groups, traits, diseases, or population structures. Combined with the robustness in handling paired-end and mate-pair data, the software shows promise to be useful for SV analyses in many bio-medical fields.
2:00 PM – 3:00 PM
DATA STANDARDIZATION AND INTEGRATION
MODERATOR: Neil Miller
Joshua Habiger | Heterogeneity and Over Dispersion in RNA-seq Data: An Empirical Bayes Approach
In the analysis of RNA-seq data, one objective is to determine which among thousands of features or genes are associated with a treatment or covariate. Recent literature has suggested that heterogeneity across features and samples, as well as over dispersion, can result in misleading and/or inefficient inference. This talk presents an empirical Bayes false discovery rate method that handles heterogeneity and over dispersion. The method is applied to 16S-rRNA data, where the goal is to identify bacteria in the microbiome of wheat that are moderately or strongly associated with productivity. It is shown that competing methods tend to identify the most abundant bacteria, even if they are only weakly associated with productivity, while the proposed method identifies more strong associations and no weak associations.
Ashiq Masood, MD | How to Make Clinical Sense and Integrate Cancer Omics Data in Clinical Practice: Challenges Moving Forward
Cancer is a complex disorder driven and governed by the genetic alterations. Studying these genetic changes on a massive scale has become feasible with the advent of Next Generation Sequencing technologies (NGS). Large scale unbiased sequencing projects such as The Cancer Genome Atlas (TCGA) have sequenced thousands of tumor samples of multiple tumor types that have substantially increased a deeper understanding of cancer initiation, progression, and metastasis. However, these sequencing efforts have created terabytes of data and to analyze these datasets we require bioinformatics skills and massive computing infrastructure. Most, if not all, clinical oncology training programs have not incorporated cancer genomics and bioinformatics in their teaching curriculum. Also, clinicians are not trained to be able to interpret cancer genomics results. This problem is further complicated by the difficulties in integrating disparate data such as cancer genomics with patient data such as electronic medical records (EMR): this requires significant infrastructure resources and experts who have the understanding of cancer genomics, computational biology, and clinical oncology. Therefore, a significant gap remains in the field of cancer genomics and its application in the clinic. We will discuss the challenges and possible solutions of cancer omics data standardization and integration to improve patient care in the era of cancer genomics.
Tomas Helikar | Cell Collective: Scalable, Reproducible, and Reusable Computational Models of Large-Scale Biological Processes
Dynamical network models of biological systems aid in data interpretation, understanding biological functions, and making predictions that can be further tested and validated in the laboratory. However, computational models involving complex mathematics and the need for computer programming have been generally limited in their utility to those with training in computational methods. Furthermore, limited reproducibility and transparency of computational models and simulated experiments makes it difficult to building on existing computational models to make them more comprehensive and accurate. To address this issue, we have developed Cell Collective, an on-line collaborative platform for the construction, annotation, simulations, and analyses of computational network models. New models and experiments in Cell Collective can be shared with the community, while models published in the platform by others can be easily further refined, extended, and built upon.
3:00 PM – 3:30 PM
3:30 PM – 4:30 PM
MODERATOR: Trupti Joshi, PhD
Bimal Balakrishnan, PhD | What Does Virtual Reality and Motion Capture Offer Healthcare?
Design and evaluation of healthcare environments have never been an easy task for architects given their complexity. Medical equipment and furniture are highly adaptable to meet a variety of use-case scenarios. Thus, capturing and evaluating affordances of human-environment interactions to evaluate a given design is a challenging task using traditional media. Increasing affordability of virtual reality (VR) and motion capture technology opens up new opportunities for rapid prototyping of healthcare environments as well as empirically evaluate human performance. Interdisciplinary work at the Immersive Visualization Lab (iLab) at the University of Missouri explores the potential of VR and motion capture technology for healthcare prototyping. These technologies also offer new opportunities for training medical professionals and present a data-driven approach to benchmark performance. This presentation will share ongoing work that explores the potential of these new technologies for healthcare.
Sejun Song, PhD| Data Visualization in the Internet of Things Era
We are at a point in society where the world around us is deeply embedded with smart things that are wirelessly connected to each other and eventually through the Internet (the Internet of Things (IoT)). IoT is attracting huge interest and will have 50 billion connected things by 2020, each being able to collect and transmit data. This huge connected possibility comes a lot of opportunities to see relationships between different ‘things’. However, the use of powerful and simple data visualizations is an essential element to present insights hidden within this avalanche of the nuanced, complex, and heterogeneous datasets. In this talk, we present a couple of our current IoT data visualization projects in the area of real-time network management and crowd science and explore the potential challenges and benefits for applying these IoT data visualization technologies to bioinformatics.
Neil Mardis | Medical 3D Printing: Expanding Services to Have a Greater Impact
The price of 3D printers has decreased markedly in recent years. High quality equipment is now readily available for use in the hospital environment. Desktop fused deposition modeling printers can successfully create clinically relevant models for surgical planning, patient and student education, orthotics/prosthetics, and research endeavors. 3D printing has positively impacted multiple clinical services at our institution including orthopedic surgery, ENT, craniofacial surgery, cardiology, cardiothoracic surgery, and rehabilitation medicine. How can we expand our services to have a greater impact on patient care and physician satisfaction?
Student Presentation: Jeremy Provance| Visualization of Molecular Structures Using HoloLens-Based Augmented Reality
Biological molecules and biologically active small molecules are complex three dimensional structures. Current flat screen monitors are limited in their ability to convey the full three dimensional characteristics of these molecules. Augmented reality devices, including the Microsoft HoloLens, offer an immersive platform to change how we interact with molecular visualizations. We describe a process to incorporate the three dimensional structures of small molecules and complex proteins into the Microsoft HoloLens using aspirin and the human leukocyte antigen (HLA) as example molecules. Small molecular structures can be introduced into the Holo Studio application, which provides native support for rotating, resizing and performing other interactions with these molecules. Larger molecules can be imported through the Unity gaming development platform and then Microsoft Visual Developer. The processes described here can be modified to import a wide variety of molecular structures into augmented reality systems and improve our comprehension of complex structural features.
4:30 PM – 5:30 PM
MODERATOR: Devin Koestler, PhD
Sarah Fouquet | Human Factors and Sociotechnical Systems Analysis
The field of Human Factors seeks to understand the relationship between people and the systems they work in which include workspaces, technology, culture, and organizations. This understanding is then applied to increase performance and decrease errors in all settings of healthcare. Utilizing a sociotechnical systems based approach to problems allows a big picture of work processes that focuses on changing systems to fit people, rather than fitting people to the system. This approach applies theory, human factors principles, and a wide array of data collection methods to improve systems and ultimately patient safety.
Jannette Berkley-Patton, PhD | Design, Evaluation, and IT Needs of Multilevel Health Promotion Interventions in African American Churches
African Americans are disproportionately affected by a number of chronic diseases including HIV, STDs, diabetes, and heart disease. Considering the high rates of church attendance and importance of religiosity among most African Americans, Black Churches can serve a key role in increasing the reach of health promotion interventions in collaboration with health agency partners. In addition, the multilevel infrastructure of many Black Churches can be seen as a naturalistic system that can be tapped for dissemination of multilevel intervention strategies. Doing so may increase intervention exposure and uptake of desired behaviors among church and community members, particularly when these strategies are delivered by faith leaders and health partners through existing church systems (e.g., ministry groups, church services, community outreach ministries). Lessons learned and future IT needs in designing and evaluating intervention dissemination processes across multilevel faith-based systems will be discussed based on current chronic disease intervention studies in African American churches.
Ruthie Angelovici | Unraveling the Genetic basis of Seed Amino Acids Composition Using Network Guided GWAS
Seeds are a primary source of proteins for both humans and livestock. However, seeds of the major staple crops such as corn and soy are deficient in several of the essential amino acids (EAA), which animals must obtain from their diet. EAA deficiency can lead to a syndrome in humans called protein-energy malnutrition, which is manifested by a series of severe symptoms including lower disease resistance, anemia, and retarded mental and physical development. The World Health Organization estimated that 30% of the population in developing countries suffers from this syndrome. To improve seed composition, one needs to gain fundamental understanding of the genetic mechanism regulating seed amino acid (AA) composition. Several studies have shown that genome-wide association mapping is an efficient approach to discover the genetic basis of several metabolites levels. However, seed AA composition is most often driven by small effect QTLs which are hard to detect using the current association panels. Our study suggests that in tightly correlated metabolic networks such as amino acids, using a combination of metabolic correlation-based network analysis and GWAS can help uncover additional genes that are part of the genetic architecture of these complex traits.
Dean Gray, PhD | Moving the Lab to the Sample: Deploying Bioinformatics Capabilities to Support Forensic and Diagnostic Analyses at the Source
Critical benefits are derived from deploying fully equipped, containerized laboratories into remote environments in response to global health concerns. The ability to provide near real-time information and intelligence at the sample source is the primary motivation, and many challenges must be overcome in order to successfully generate and deliver defensible data. Deploying bioinformatics capabilities is a relatively new development driven by customer’s needs across many technical areas, and is supported by innovations in rugged, miniaturized and transportable equipment. Considerations such as power source, dependability, logistics, sample integrity, and data capture with recent examples and applications to One Health Intelligence will be discussed.
5:30 PM – 6:30 PM
COCKTAILS & NETWORKING | Room 401
6:00 PM – 8:00PM
DINNER | Room 401
FRIDAY, APRIL 14, 2017 | UMKC PIERSON AUDITORIUM
7:15 AM–7:45 AM
7:45 AM–8:00 AM
WELCOME & INTRODUCTIONS
Keith Gary, PhD | Thank Sponsors and Volunteers
Mark Hoffman, PhD | Summary of Agenda and Introduce Keynote Speaker
8:00 AM – 9:00 AM
Philip Payne, PhD, FACMI | Precision Medicine and Healthcare Transformation: A Tale of Two Populations
The healthcare and life science research, education, and practice domains are undergoing tremendous change, driven by a combination of economic, policy, and technology factors. At the core of these changes are: 1) an increased focus on the use of patient-derived data, contextualized by the best available scientific evidence, to deliver highly tailored and individualized wellness and clinical care interventions; and 2) a simultaneous demand to transform healthcare at a population-level through the pursuit of a triple aim concerned with increasing quality and safety while also decreasing costs. While these two areas of endeavor are often approached as distinct from each other, they in fact share a common set of data, information, and knowledge based needs that can be satisfied using a systems-level approach to biomedical informatics and data science. This presentation will explore these critical issues, focusing upon the changing landscape of healthcare and life science, the role of computation and data science in both driving and adapting to such change, and the enumeration of currently open research, development, and innovation opportunities that can and should be pursued in order to improve the health of individuals, their families, and their communities-at-large
9:00 AM – 10:00 AM
MODERATOR: Julie Banderas, PharmD
Jenifer Allsworth, PhD | Investigating Adverse Peripartum Outcomes using Electronic Health Record Information
Adverse pregnancy outcomes at the time of delivery, including postpartum cardiomyopathy and hysterectomy are uncommon, but often serious. This talk will share examples of the development of a pregnancy cohort and evaluation of racial disparities in peripartum adverse outcomes using electronic health record information.
Janelle Noel-MacDonnell, PhD | Utilization of Multi-level data to Forecast Asthma Related Emergency Department Visits
We are at a point in time where the amount of data available has drastically outpaced its utilization potential. This current wealth of data has also prompted the research community to creatively tackle many of their research questions in different ways, especially for those questions regarding healthcare. In this talk, Dr. Noel-MacDonnell will discuss some of her recent work that involves combining multiple levels of data to develop a forecast model to predict the number of emergency department (ED) visits attributed to asthma in the pediatric population. Her approach uses data, collected weekly, from the Pediatric Health Information System (PHIS) and the U.S. Environmental Protection Agency (EPA).
Dong Xu, PhD | Application of Deep Learning in Studying Proteins
Deep learning, as the cutting-edge technology in machine learning, has produced remarkable improvements on various computational problems and presents a new opportunity for studying proteins at the sequence and structure levels. The growing amount of various protein data also allows deep learning to generate robust models. We have demonstrated success on applications of deep learning in several protein bioinformatics problems, including protein structure prediction, protein localization prediction, and protein post-translational modification site prediction. Our long-term goal is to develop a general deep-learning framework for many other protein analysis and prediction problems.
Student Presentation: Katherine Shortt| Identification of Novel Regulators of Acetaminophen Induced Hepatocyte Toxicity Using a Genome Scale CRISPR-Cas9 Screening System
Acetaminophen(APAP) is a commonly used analgesic responsible for over 56,000 overdose-related emergency room visits annually. Due to a long asymptomatic period and limited treatment options, there is a high rate of liver failure resulting in either organ transplant or mortality. The underlying molecular mechanisms of injury are not well understood and effective therapy is limited. We hypothesize that the mechanism of APAP toxicity is more complex than is currently known. Identification of new genetic risk factors would provide new mechanisitic insights and new therapeutic targets for APAP induced hepatocyte toxicity or liver injury. This study uses a genome-wide CRISPR-Cas9 screen to evaluate genes that are protective or susceptible to APAP-induced liver injury. HuH7 human hepatocarcinoma cells expressing Cas9 were transduced with a GeCKOv2 sgRNA library targeting over 20,000 genes. Cells containing gene knockouts were treated with 15mM APAP for 30 minutes to 4 days. A gene expression profile was developed based on the 1)top screening hits, 2)overlap expression data from human microarrays of APAP overdose patients, and 3)prediction of affected biological pathways. This screen is the first genome-wide CRISPR-Cas9 knockout screen of APAP-induced hepatocyte toxicity. The top screening hits include novel genes previously not linked to liver injury. We further demonstrate the implementation of intermediate time points for the identification of early and late response genes. A negative selection screen identifies genes involved in fundamental processes, including NAAA, C19ORF60, and MYOZ3. A positive selection screen identifies numerous genes potentially involved in pathogenic processes, including LZTR1, PGM5, and KIF23. A top essential pathway at 24 hours of APAP treatment is Regulation of Skeletal Muscle Contraction. Collectively, this study has illustrated the power of a genome-wide CRISPR-Cas9 screen to systematically identify novel genes involved in APAP induced hepatocyte toxicity and to provide potential new targets to develop novel therapeutic modalities.
10:00 AM – 10:30 AM
10:30 AM–11:30 AM
MODERATOR: Lawrence Dreyfus, PhD
Sarah Dallas, PhD | Live Cell and Intravital Imaging and the Challenges of Data Handling and Interrogation
Live cell and intravital imaging allows us to view biological events from a dynamic perspective and provides novel insights into biological processes that cannot be obtained by viewing them using static (snapshot) approaches. Our laboratory employs live cell and intravital imaging approaches to investigate bone cell biology and physiology, including understanding the dynamic properties and interactions of bone cells and the dynamic processes of extracellular matrix assembly and mineralization in bone. A challenge of these studies is the large amount of image data that they generate. For example one of our typical live imaging experiments may generate ~50,000 images, equating to ~30-50Gb data. These timelapse image stacks are data rich and can be mined and interrogated to generate quantitative outputs that quantify cell behaviors. Once archived, the data can be re-interrogated in the future to address new biological questions that arise based on advanced understanding of the biology. Archiving this data and efficiently mining the data to generate quantitative outputs are major challenges in the field that could greatly benefit from the application of new and creative bioinformatics approaches.
Michael Veeman | Whole-organ Cell Segmentation in the Ciona Notochord
My lab combines functional genomics and quantitative imaging approaches to work towards a systems level understanding of chordate morphogenesis. Our model organism is the simple chordate Ciona, which has a stereotyped chordate body plan in an embryo small enough to easily image in toto with fine subcellular detail in a single field of view. We have developed a set of computational tools for semiautomatically segmenting all 40 notochord cells in 3D to quantify diverse aspects of cell size, shape and tissue architecture. I’ll discuss our segmentation and analysis workflow, which is based on the seeded watershed transform, and briefly discuss the biological insights that have followed from extensive analysis of notochord cell size and shape
Susan Brown, PhD | Whole Genome Mapping
Restriction maps where once considered indispensable genomic resources and still provide powerful means of validating genome assembly, extending contiguity, and analyzing structural variation. However, conventional restriction mapping methods are tedious, time-consuming and cost prohibitive to all except the most heavily supported model systems. We provide access to the optical mapping methodology developed by BioNano Genomics to generate whole genome restriction maps from ultra long DNA molecules. This next generation mapping technology is efficient and affordable. I will discuss the application of BNG next gen mapping to validate and improve draft genome assemblies.
Student Presentation: Cynthia Chen| Automated Secondary Analysis Pipeline for RNA-seq and ChIP-seq Data
Along with the NGS technology improvement, RNA-seq and ChiP-seq sequencing data is now generated on a daily basis. It’s crucial to quantify data quality before starting further downstream analysis. NGS analysis results often get misinterpreted due to lack of sufficient preliminary quality control. We developed an automated pipeline that implements various current quality control tools and generates an interactive html report using R markdown. For RNA-seq datasets, we measure alignment statistics, feature distribution, coverage plot, strand specificity, FPKM distribution, ERCC count table, among others. For ChIP-seq datasets, we assess alignment statistics, create a cross correlation plot, measure IP efficiency, and create bigWig track files for visualization. Using the report from our pipeline, researchers can easily quantify their sample quality before rushing into further analysis. The user-friendly interactive web report is also easy to compare between experiments and share with lab members. The entire pipeline is available in GitHub and can be custom built for different needs.
11:30 AM–12:30 PM
11:30 AM–11:35 AM
MODERATOR: Donna M. Buchanan, PhD
Trupti Joshi, PhD | Addressing Precision Agriculture and Food Security via Informatics Approaches
With the world population projected to cross 9.7 billion by the year 2050, concrete steps are required to increase the food production to match the demand. Advancement in next generation sequencing (NGS) and high-throughput technologies has resulted in generation of ‘MULTI-OMICS’ data for many crops. However, these data often remain scattered across different repositories, making it difficult to integrate and mine them. Informatics plays a crucial role in providing solutions to address these challenges, mine crop germplasm genomic resources, facilitate efficient data sharing and identify opportunities for improved crop production. We have achieved this by developing successful solutions such as Soybean Knowledge Base (SoyKB) and Knowledge Base Commons (KBCommons) frameworks equipped with tools for data integration, analytics, visualization and interpretation for precision agriculture.
Kim Smolderen, PhD | Implementing Personalized, Evidence-Based Tools Predicting Patients Outcomes to Make More Informed, Shared Decisions for Peripheral Arterial Disease Treatment
The primary treatment goals for peripheral arterial disease (PAD) – atherosclerotic disease in the lower-extremity arteries – are symptom relief, quality of life improvement, and cardiovascular risk reduction. Several treatment options are available for PAD, ranging from invasive revascularization procedures to non-invasive options, including exercise therapy, PAD-specific medications, and cardiovascular risk management.15 In clinical scenarios where there is no “gold-standard” treatment, like in PAD, and a rapidly growing market for newly introduced technologies, including medical devices for invasive PAD procedures (e.g. stents for endovascular treatment), with limited performance measurement and accountability criteria, there is a high risk of unwanted variation in treatment practices, misallocation of treatments, and unnecessary costs. Given this context, some of the current challenges in current PAD care include: 1) limited access to the evidence- base in routine clinical care for patients and providers; 2) the potential mismatch of PAD treatments to patient preferences and profiles; and 3) patients not being informed or engaged in medical decision making. These challenges may leave patients uninformed about treatment risks and benefits, increase the risk of misallocating treatments to patients, and may unnecessarily increase costs. A very promising strategy to overcome these challenges is the use of evidence-based, personalized decision support tools.
Sarah Soden, MD | Big Data – Small Patients
The Center for Pediatric Genomic Medicine at Children’s Mercy utilizes sequencers that are capable of producing a terabyte of data in a single run, yet the etiology of a patient’s symptoms may be as subtle as a change in a single nucleotide. Extracting accurate diagnoses from voluminous genomic datasets requires a team with combined expertise in molecular biology, bioinformatics, and human genetics. The ability to simultaneously sequence and analyze all genes that may be causing a patient’s symptoms is particularly beneficial for diagnosing patients with very rare disorders, newly discovered genetic diseases, and those with atypical presentations. Bringing a child’s diagnostic odyssey to an end is often a powerful moment for families and, in some cases, can affect medical care and even survival.
Janakiraman Subramanian | Translating Cancer ‘Omics’ Data to Improve Patient Outcomes
Next generation sequencing (NGS) technologies have revolutionized the field of cancer ‘omics’. NGS allows for rapid and comprehensive sequencing of large numbers of tumor samples. This has allowed large scale sequencing efforts such as The Cancer Genome Atlas (TCGA) and others to successfully describe the molecular landscape of a wide range of cancers. In selected tumor types this information has been effectively used to improve patient outcomes. But the overall pace of progress has been slow and despite the promise of NGS, oncologists are faced with the day to day challenge of patients with cancer who are running out of treatment options. There is a need to develop comprehensive clinical programs that integrate NGS in the management of patients diagnosed with cancer. At the Saint Lukes Cancer Institute (SLCI) we have put forward a vision of integrating cancer “omics’ driven precision medicine in the oncology clinic. It will be a collaborative effort between the SLCI, industry and payers, it would support oncologists in identifying the best treatment options using NGS based tumor assay. We will be discussing specific patient examples to illustrate the value of this model in bringing novel therapies to the bedside and also illustrate the continuing challenges in integrating precision medicine in the clinic.