Plenary Talks
Venue, Hotels and Social
Paper Submission

Invited talks and Tutorials

Invited talks have always been an important way to add value to the conference, not only for the presentations but also because of the interactions during the event.

This webpage will be updated as invited speakers agree to give talks and tutorials.

Invited talks

Richard B. Berry, PhD, Professor of Medicine, Medical Director, UF Health Sleep Disorders Center, Sleep Medicine Fellowship Director
Biomedical Computation - Sleep Medicine Opportunities

Sleep studies acquire a tremendous amount of digital data but currently the methods of analysis are crude and labor intensive. There are numerous opportunities for the development of computational tools to improve our understanding of human sleep and disorders such as sleep apnea. Several areas where biomedical computational techniques would be of great interest are discussed including sleep staging, automated scoring of sleep studies, detection of arousals, improved analysis of the oximetry signal, and estimates of sleep depth will be reviewed.

Ana Conesa, PhD, Head of Genomics of Gene Expression Lab, University of Florida
Functional transcriptomics in the post-NGS era: multiomics integration and new technologies

Next generation sequencing has speed up genome analysis and brought omics research closer to many organisms and biological scenarios. Today an increasing number of research projects propose the combined use of different omics platforms to investigate diverse aspects of genome and transcriptome function. However, combination of high-throughput data from different technologies in not straight forwards and it is not clear what the best way is to combine omics information to achieve interpretable muti-layer systems biology models. On the other hand, short read sequencing has shown to have intrinsic limitations to accurately describe the molecular and functional complexity of the transcriptomes of higher eukaryotes. I will present novel computational approaches in my lab to integrate omics technologies and use third generation sequencing platforms to deepen in the architecture, regulation and functionality of gene expression.

Rebecka J├Ârnsten, PhD, Associate Professor, Chalmers University
Network modeling of TCGA data: integrative and disease comparative approaches

Statistical network modeling techniques have the potential to increase our understanding of cancer genomics data. Here, we analyze multiple TCGA data sets via a generalized sparse inverse covariance model, carefully addressing such challenges as unbalanced sample sizes, local network topology, model selection and robust estimation. The method integrates genetic, epigenetic and transcriptional data from multiple cancers, to define links that are present in multiple cancers, a subset of cancers, or a single cancer. The modeling results are available at cancerlandscapes.org, where the derived networks can be explored as interactive web content and be compared with several pathway and pharmacological databases. Network components are shown to fall in mainly two categories: common to all cancers or unique to one type of cancer. We also discuss how network models can be used to construct diagnostic markers (predictors of survival). This is joint work with the Nelander lab, SciLife, Uppsala.

Xiaolin Li, PhD, Associate Professor, University of Florida
DeepHealth: Deep Learning and its Applications in Precision Medicine

In recent years, Deep Learning (DL) has attracted tremendous enthusiasm in both academia and industry, winning numerous competitions in computer vision, natural language processing, and speech recognition. The key breakthrough in DL was due to a series of improvements in artificial neural network, machine learning, big data, and big systems. Researchers have developed large-scale deep learning models with billion to trillion parameters on tens thousands of CPU cores and GPUs. In this tutorial, we will give a broad overview of the state-of-the-art in deep learning: deep neural networks, deep convolutionary neural networks, deep recurrent neural networks, and enhanced models. We will also present representative applications in personalized precision medicine: sensing, monitoring, diagnosis, and prediction for bipolar, cancers, sepsis, and others.

Giuseppe Nicosia, PhD, Associate Professor, University of Catania
Multi-Objective Multiplex Multi-Omic Genome-Scale Models for Cancer Metabolism

Cancer cells were found to have a specific metabolism that is remarkably different from the tissues from which they originated, due to their high demand for proteins, lipids, nucleotides and energy levels, all necessary for speed-up and enhanced growth and proliferation.
The metabolic network is highly interconnected and complex, and probably the best characterized biological network in terms of models and multi-scale omic data. This help us to study the global goals and functional implications of the dysregulated metabolism in cancer.
In this talk he will show the effective integration of different sets of omic datasets, multi-objective optimization, multiplex and multi-scale genome-scale models to try to decipher the tumor metabolism.

Jose C. Principe, PhD, Distinguished Professor of Electrical Engineering, University of Florida
A Transient Model for Neuro-Modulation

This talk proposes a novel approach to quantify brain activity taking into consideration the transient nature of the electroencephalogram (EEG) and local field potentials (LFP). In fact, the quantitative structure in the EEG/LFP extracted visually by clinicians relates to transients called bursts or spindles at selected frequencies called the EEG rhythms. Our framework proposes a noisy combination of filtered shot noise models that extract these events from a single channel of EEG/LFP using Matching Pursuit. The filters are defined a priori in a dictionary that spans the event shapes. Conceptually, the filters can also be learned from the data. We present results from rat hippocampal and human EEG data to illustrate the methodology.

My T. Thai, PhD, Professor, University of Florida
Group Testing and Its Application in Biological Screens.

Group testing has various applications in blood testing, chemical leakage testing, coding, multi-access channel communication, and many others. In the context of biology, group testing is usually referred as pooling designs. As the technology for obtaining sequenced genome data is getting mature, more and more sequenced genome data are available to scientific research community, so that the study of gene functions has become a popular research direction. Such a study is supported by a high quality DNA library which usually is obtained through a large amount of testing and screening. Therefore, the efficiency of testing and screening becomes very important. Pooling design is a tool to reduce the number of tests in DNA library screening as well as in DNA microarrays. The construction of pooling design and its ability for decoding are very challenging, especially when inhibitors are present in the biological sample. Towards this end, we are tackling this problem in several layers, from distinguishing positive clones from negative clones to a more complicated model, considering errors and inhibitors.

Madhav Thambisetty, MD, PhD, Chief of Clinical and Translational Neuroscience, NIH
Seeking biomarkers and understanding mechanisms: building an integrated approach to Alzheimer's disease.

We have pursued numerous 'omics'-based approaches to identify predictive biomarkers of Alzheimer's disease (AD). In combination with neuroimaging, these methods have been successfully used to discover blood biomarkers reflecting core pathological features of AD. We have recently identified Alpha2 Macroglobulin (A2M) as a candidate serum biomarker predictive of incident AD in cognitively normal individuals. A2M is also associated with cerebrospinal fluid changes related to AD pathology. Using a network analysis, we then identified regulators of A2M gene expression and examined the relationship between peripheral and central A2M gene expression. Regulator of Calcineurin (RCAN1) drives A2M gene expression and inhibits calcineurin, a major brain tau phosphatase. Finally, using label free proteomics, we demonstrate that protein levels of A2M and calcineurin in the brain are positively correlated. These studies identify a novel molecular pathway linking the acute phase protein, A2M and neuronal injury in AD.