| TIME |
TITLE |
| 7:30 |
Welcome and Overview - Mike Hawrylycz |
| 7:35-8:00 |
Andrew Su
Laboratory of Neuroimaging, Los Angeles, CA
Using expression QTLs to reveal tissue-specific networks.
Abstract:
Through our SymAtlas data set and web site, we have used gene expression analysis to reveal differences in transcript levels across a diverse set of tissues in a single genetic background. Recently, we have expanded this effort in a second dimension, tracking changes in expression in select tissues across many genetic backgrounds. By combining these data with a dense genotype map, we have used measurements as phenotypes in whole-genome association studies. The resulting "expression QTL" maps reveal many insights into transcriptional regulatory networks, and allow description of differences between tissues at the level of biological pathways.
Biosketch:
Since joining GNF in 2002, Andrew Su has been the head of the Computational Discovery group. One focus of his research is gene annotation and transcript validation. His group also maintains several collaborations with experimental scientists studying a variety of biomedical topics.
URL:
web.gnf.org/scientific/informatics/compdis.html |
| 8:00-8:25 |
Desmond Smith
UCLA School of Medicine, Los Angeles, CA
The brain cubed: mapping transcripts and proteins in 3D
Abstract:
We are creating 3D molecular maps of the brain using a method called voxelation. In this method, spatially registered voxels (cubes) are analyzed using microarrays or mass spectrometry. We then employ the mathematical methods of biomedical imaging to reconstruct transcript and protein maps at a genome-wide scale.
Biosketch:
Dr. Desmond Smith obtained his MA and MD from the University of Oxford and his PhD from the University of Cambridge. After postdoctoral training at Harvard and Berkeley he obtained a faculty position at UCLA, where he is currently Associate Professor of Medical and Medical Pharmacology.
URL:
labs.pharmacology.ucla.edu/smithlab/ |
| 8:25-8:45 |
Coffee Break |
| 8:45-9:10 |
Paul Pavlidis
Columbia University Medical Center, New York, NY
Scaling up microarray informatics for the brain
Abstract:
Thousands of gene expression microarray experiments have been done, including many of relevance to neuroscientists, but there are few ways to efficiently mine this accumulation of data. I will describe work we have been doing to develop a large-scale database and analysis system for microarray data (Gemma) and our efforts to identify patterns of gene expression that are reproducible between labs. I will present data indicating that comparing and combining results across laboratories is a fruitful endeavor that can result in the identification of novel and interesting expression patterns to neuroscience.
Biosketch:
Dr. Paul Pavlidis received his PhD in molecular and cell biology / neuroscience from the University of California, Berkeley in 1994. After completing post doctoral studies as an electrophysiologist at Stanford (laboratory of Dan Madison) and Columbia (jointly in the labs of Eric Kandel and Steven Seigelbaum) he began work in bioinformatics with William Noble at Columbia in 2000. He joined the Columbia faculty in Biomedical Informatics in 2003, with appointments in the Columbia Genome Center and the Center for Computational Biology and Bioinformatics.
URL:
microarray.cu-genome.org |
| 9:10-9:35 |
Amir Assadi
University of Wisconsin, Madison, WI
Applications of the Mathematical Theory of Complex Dynamical Systems to Gene Expression Data Analysis
Abstract:
Dynamical systems theory has been widely applied in models of cells and networks in neuroscience. This work presents an extension of related concepts to neurogenomics with focus of the lecture on design of novel algorithms that complement the methods from statistical genomics. Some preliminary case studies indicate that the proposed algorithms allow a hierarchical clustering of microarray data that noticeably improves upon the common statistical approaches. Further, the biological intuition for plausibility of our algorithm eliminates the need for making ad hoc statistical assumptions. Examples from recent microarray data will be presented.
Biosketch:
Amir Assadi is a mathematics professor at UW Madison and an affiliate member of The Wisconsin Genome Center. He received his undergraduate degree from UC Berkeley, his MS and PhD from Princeton and also is an alumnus of the MBL at Woods Hole for training in computational neuroscience and neuroinformatics. His current research interests focus on analysis of data, design of data-driven models for brain functions, especially in human vision and pain. More recently, his research includes dynamic modeling of neuronal events at subcellular and molecular scales, the relevant neurochemistry and molecular biology.
URL:
www.math.wisc.edu/~assadi |
| 9:35-10:00 |
Jonathan D Victor
Weill Medical College of Cornell U., New York, NY
Metric-space analysis of neural activity patterns: sequence analysis in time
Abstract:
The biophysical basis of neural activity is well understood, but the relationship between the pattern of firing of neurons and behavior (percepts, actions, intentions) remains a central problem in neuroscience. Probing this relationship requires identification of statistical correspondences between patterns of neural activity and their behavioral correlates. I describe a theoretical framework for addressing this abstract problem, and application of this framework to recordings of single- and multi-neuronal activity in the visual cortex of the macaque. The key algorithmic steps are generalizations of the dynamic programming algorithms of Needleman-Wunsch and Sellers in common use for comparison of genetic sequences.
Biosketch:
Dr. Jonathan Victor received his MA in mathematics from Harvard College, Cambridge, MA, PhD from The Rockefeller University, New York, NY and MD from Cornell University Medical College, New York, NY. He is currently professor of Physiology, Biophysics, and Systems Biology, Weill Graduate School of Medical Sciences, professor of Department of Neurology and Neuroscience and Weill Graduate School of Medical Sciences and attending Neurologist and Head, Division of Clinical Neurophysiology, Department of Neurology and Neuroscience, The New York Hospital - Cornell Medical Center.
URL:
Further background and references related to this talk:
www-users.med.cornell.edu/~jdvicto/metricdf.html
Key review article:
www-users.med.cornell.edu/~jdvicto/vict05.html
Laboratory of Neuroinformatics Home Page
neurodatabase.org/index.html
Speaker's Home Page
www-users.med.cornell.edu/~jdvicto/jdvonweb.html |
| 10:00-10:30 |
Panel Mediated discussion Session |
| 10:30-15:30 |
Break |
| 15:30-15:55 |
Giorgio Ascoli
The Krasnow Institute, Geroge Mason University, Fairfax, VA
Three dimensional reconstruction of the rat hippocampus.
Abstract:
The hippocampus is implicated in important cognitive functions, including the consolidation of autobiographic memories in humans, and spatial navigation in rodents. We have quantitatively integrated the cellular and system level neuroanatomical representations of the rat hippocampus by creating a virtual reality model based on high-resolution experimental data. Starting from ex-vivo microscopic MRI and cryostatic Nissl stain, 3D surfaces of internal and external cytoarchitectonic boundaries of the hippocampus, and the respective volumes, are rendered with a multi-step computational process. Individual neuronal reconstructions are systematically added and oriented in the appropriate layers. The model, validated with stereological data, is suitable to estimate the subcellular distribution of expression patterns and potential synaptic connectivity.
Biosketch:
Dr. Giorgio A. Ascoli received a Ph.D. in Biochemistry and Neuroscience from the Scuola Normale Superiore of Pisa, Italy, and continued his research at the National Institutes of Health in Bethesda, MD. Dr Ascoli moved to the Krasnow Institute for Advanced Study at George Mason University in 1997, and is the founder and head of the Computational Neuroanatomy Group, a multidisciplinary research team which includes psychologists, biologists, physicists, computer scientists, engineers, mathematicians, and physicians.
URL:
krasnow.gmu.edu/L-Neuron |
| 15:55-16:20 |
David Craig
TGen Instittute, Phoenix, AZ
NIH Neuroscience Consortium: Leveraging neuroinformatics and the human genome project into a single data structure.
Abstract:
Rapid acquisition and dissemination of high quality genomic data in a single web accessible data structure is essential to moving translational research forward to the clinic. Resulting from the Neuroscience Blueprint, 15 Institutes of the National Institutes of Health have established the NIH Neuroscience Microarray Consortium to provide NIH funded neuroscience investigators with the use of gene expression profiling technology and SNP genotyping. A major accomplishment of this consortium is development of a single web-accessible data structure that includes project annotation in MAGE-ML, image files, linked ontology and high-resolution phenotypic information. All raw data, including project annotation, generated by the consortium is publicly available six months. In this presentation, we will describe development and application of this neuroinformatic tool. We seek to integrate region and cell specific expression signatures with temporal/special neuroanatomic maps to ultimately be enabled to understand the neural circuitry responsible for the brain's cortex functions.
Biosketch:
Dr. David Craig is an associate investigator and faculty member at The Translational Genomic Research Institute (TGen). Dr. Craig completed his post-doctoral training at TGen in Neurogenomics, completed his Ph.D. at the University of Washington in Computational bioengineering, and completed his B.S. at the University of Arizona. Dr. Craig's research interests include identification of predictive genetic signatures of disease using gene expression and SNP genotyping microarrays.
URL:
arrayconsortium.tgen.org |
| 16:20-16:45 |
James Gee
UPenn, Philadelphia, PA
Challenges in large-scale gene expression image analysis
Abstract:
Toward the creation of highly detailed maps of gene expression, a natural means of integrating and organizing the information is to establish a referential coordinate system with which the data can be indexed. Such a coordinate system would make possible localization and comparison of gene expression information from, for example, in situ hybridization. The resultant atlases implement a flexible means for storing and accessing image-based genomic and phenotypic information, enable model-based methodologies to extract the latter, and facilitate genotype-phenotype correlative studies. In this talk, we describe some of the challenges involved in 1) developing coordinate systems that capture in detail the volumetric in vivo structure of the model animal of interest based on highly resolved histologic acquisitions from an exemplar individual, and 2) populating an atlas with gene expression data from a very large number of experimental animals.
Biosketch:
James Gee, Ph.D., is Associate Professor of Radiologic Science and Computer and Information Science at the University of Pennsylvania. He received the B.S. degrees in Computer Science and Electrical Engineering from the University of Washington, Seattle, and the M.S. degree in Electrical Engineering from the same institution. He holds a Ph.D. in Computer and Information Science from the University of Pennsylvania, where he was awarded an R01 grant on medical image analysis while still pursuing graduate studies. Best known for contributions to computational anatomy, his research interests include biomedical imaging, probabilistic and geometric modeling, pattern analysis, and scientific computing.
URL:
www.grasp.upenn.edu/biomedical |
| 16:45-17:00 |
Coffee Break |
| 17:00-17:25 |
Sayan Pathak
Allen Institute for Brain Science, Seattle, WA
Clustering gene expression profiles in mouse brain
Abstract:
Allan Brain Altas (ABA) currently comprises of high resolution brain images of over 12000 genes in adult C57BL/B6 mouse. Analysis of such large scale gene expression data in high resolution brain images opens new doors for a refined interpretation of neuroanatomy. Automated atlas based registration technology is used to map mouse brain gene expression to a common anatomical framework. The resultant anatomic mapping is used to calculate expression profiles for different brain structures. In this talk, we will present a means to effectively search for gene expression in our ABA image database and present our initial attempts on clustering the expression data. Our goal would be to lay down the foundations for a potential discussion on how the various data modeling and analysis methodologies can be applied towards refining our understanding of the neuroanatomy using on the gene expression image informatics.
Biosketch:
Sayan Pathak, Ph.D., is a senior analyst at the Allen Institute for Brain Sciences, and affiliate Assistant Professor of Electrical Engineering at the University of Washington. He received his B.S degree in Instrumentation Engineering from the Indian Institute of Technology, Kharagpur, MS and PhD degree in Bioengineering from the University of Washington. He was a key architect of the Allen Brain Atlas gene expression search methodology. His research interests include human computer interfacing, pattern recognition, and image informatics.
URL:
faculty.washington.edu/sayan |
| 17:25-17:50 |
Teiichi Furuichi
Riken Brain Science Institute, Wako, Japan
Cerebellar circuit development and gene expression
Abstract:
Mouse cerebellum develops through a series of cytogenetic and morphogenetic events (cell proliferation and migration, dendrogenesis and axogenesis, synaptogenesis, myelination, etc.) that are genetically coded within the first three weeks of life. To decipher the genetic basis for cerebellar development, we investigate the spatio-temporal gene expression profiles on a genome-wide basis (differential display, RT-PCR, GeneChip, and in situ hybridization) and informatively systematize all the profiles in a Cerebellar Development Transcriptome Database (CDT-DB). We have demonstrated that the postnatal development of mouse cerebellum is genetically programmed by thousands of genes that exhibit differential expression patterns in time and space.
Biosketch:
Dr. Furuichi is Head of the Laboratory for Molecular Neurogenesis at the RIKEN Brain Science Institute. Previously, he discovered mysterious RNA-linked DNAs and reverse transcription from bacteria (PhD study at SUNY Stony Brook), succeeded in cloning the mouse IP3 receptor cDNA (postdoc. and research assoc. at NIBB, Okazaki), and studied IP3-Ca2+ signaling (associate prof. at Univ. of Tokyo). He now directs the cerebellar development transcriptome database (CDT-DB) project, a step in deciphering the genetic blueprint of cerebellar development.
URL:
Database:
www.cdtdb.brain.riken.jp
Laboratory:
www.brain.riken.go.jp/labs/lmn/index.html |
| 17:50-18:28 |
Panel Mediated Discussion Session |
| 18:28-18:30 |
Wrap Up |
|
|
