Kenneth Lamar Lange - US grants

[unreadable] DESCRIPTION (provided by applicant): One of the paradoxes of modern genetics is the contrast between the tremendous technological advances in sequencing and genotyping during the past decade and the slow progress in identifying genes for complex diseases. These diseases involve subtle disruptions of biochemical and developmental pathways and display substantial genetic heterogeneity and gene-by-gene and gene-by-environment interactions. In response to these challenges, geneticists are collecting much larger samples and genotyping enormous numbers of SNPs (single nucleotide polymorphisms). To handle the massive increases in data flow and extract the maximum amount of information from available data, better statistical analysis tools must be made available to the human genetics community. The current grant supports construction of new statistical methods and their translation into user friendly software via the widely distributed program Mendel. Under the auspices of the grant, we will tackle a series of related projects on computational statistics, association mapping, estimation of DNA copy numbers, population genetics, and software for managing and displaying human pedigree data. [unreadable] [unreadable] Our research in computational statistics revolves around three classes of optimization algorithms - MM and EM algorithms, block relaxation methods, and lasso penalized estimation. We will apply these methods to estimation in random graphs, nonnegative matrix factorization, and multicategory discriminant analysis. These methods are also pertinent to fast logistic regression with case-control data and fast mapping of QTLs (quantitative trait loci). We further plan to develop fast tests of association based on contingency tables, robust testing procedures for multivariate traits, and algorithms for modeling gene-by-gene and gene-by-environment interactions. [unreadable] [unreadable] Our efforts on copy number variation will focus on penalized estimation of DNA copy number by signal intensity, and hidden Markov modeling of copy numbers from the Illumina genotyping platform. In population genetics we will develop methods and software for testing Hardy-Weinberg equilibrium in pedigree data, penalized estimation of haplotype frequencies, and estimation of ethnic admixture. Finally our software development efforts will concentrate on making Mendel more conducive to dense, genome-wide SNP data, including: parallelization of the existing Mendel code; restructuring of the data structures in Mendel; making it easier to run complete analysis routines within Mendel; and perfection of MendelPro, the graphical user interface to Mendel. [unreadable] [unreadable] This ambitious agenda is all part of our coherent effort to provide a single platform for managing, displaying, and analyzing genetic data. This kind of software infrastructure is necessary if genetic epidemiology is to move rapidly forward in the twenty-first century. [unreadable] [unreadable] [unreadable]

9987641
Fred Fox - University of California at Los Angeles
IGERT: Training Program in Bioinformatics

This Integrative Graduate Education and Research Training (IGERT) award supports the establishment of a multidisciplinary graduate training program of education and research in bioinformatics. As part of a major new bioinformatics initiative at UCLA, the IGERT bioinformatics program will provide rigorous training for Ph.D. candidates in thirteen participating departments, supporting them with tools for creating robust new bioinformatics applications. It also will engage undergraduates from UCLA and feeder institutions in summer research internships and support accelerated bioinformatics training that can bypass the baccalaureate and proceed directly from undergraduate admittance to an advanced degree. IGERT trainees will satisfy all major subject area requirements of their Ph.D.-granting departments and also the core components of the newly established, certificate-granting interdepartmental program in bioinformatics. This core consists of courses in genomics and bioinformatics, statistical methods in computational biology, and a weekly seminar in which students and faculty discuss specific examples of how biological problems map to are solved by approaches from multiple disciplines. The interdepartmental program also requires a subject area minor in molecular life science for trainees in computer science, mathematics and statistics; trainees in life sciences must satisfy a minor in mathematics-statistics or computer science. IGERT trainees will work in matrix environments that force regular, collaborative contacts; these are provided by the Keck Foundation-supported Bioinformatics User Centers adjacent to the UCLA-DOE Macromolecular Structure Laboratories and the Human Genetics core DNA sequencing and microarray facilities. The User Centers can be made available to visitors who wish to apply powerful, integrated bioinformatics approaches in their research. The UCLA-IGERT bioinformatics program will develop a proactive industry liaison that contributes to programmatic vision, helps test the preparedness of trainees in private sector challenges and provides networking support. The IGERT program will also collaborate with the UCLA-based, NSF-sponsored national Institute for Pure and Applied Mathematics in seminars and conferences that help define the evolution of bioinformatics as a fundamental discipline.

IGERT is an NSF-wide program intended to meet the challenges of educating Ph.D. scientists and engineers with the multidisciplinary backgrounds and the technical, professional, and personal skills needed for the career demands of the future. The program is intended to catalyze a cultural change in graduate education by establishing new, innovative models for graduate education and training in a fertile environment for collaborative research that transcends traditional disciplinary boundaries. In the third year of the program, awards are being made to nineteen institutions for programs that collectively span all areas of science and engineering supported by NSF. The intellectual foci of this specific award reside in the Directorates for Biological Sciences, Computer and Information Science and Engineering, Mathematical and Physical Sciences, and Education and Human Resources.

DESCRIPTION (provided by applicant): Turning the data collected by the recent quantum leap in DNA sequencing technology into useful biomedical knowledge is one of the greatest scientific challenges of the 21st Century. The enormous volume of DNA data now collected cannot be accessed, manipulated, or analyzed without computers. However, the greatest bottlenecks involve software rather than hardware. In spite of current levels of sophistication, we still lack adequate theory and algorithms to perform many fundamental tasks in molecular genetics and genetic epidemiology with speed and precision. It will take a new generation of scientists trained in both the biological and mathematical sciences to push forward our nation's genomic agenda. Few universities have the infrastructure and human resources to mount a genomic analysis training program of the scope possible at UCLA. Our training program's core curriculum includes work in molecular biology, human genetics, probability and statistics, bioinformatics, and biomedical ethics. Students must also prepare for and attend pertinent weekly scientific seminars. The trainees collectively meet on a monthly basis with the program faculty and directors to learn about career matters not covered in regular coursework. Each year we will train 12 predoctoral students drawn preferentially from students in the first half of their doctoral programs. Each trainee will be funded for two years contingent on satisfactory progress. In exceptional cases a third year of funding will be considered. This training grant has brought together faculty from four schools and 14 PhD-granting departments within the university to address the current gap in scientific training. During the years 2002 to 2010, this training grant has successfully completed training of 22 predoctoral students in the art of genomic analysis and interpretation. These doctorates have gone on to productive careers in genomic sciences in industry, government, and academia.

Abstract The biomedical sciences are drowning in big data. Progress in ?elds such as genomics and medical imaging is being stymied by the lack of ap- propriate computational tools. This grant promotes the development of algorithms, statistical methods, and software for the analysis of the big datasets encountered in the biomedical sciences. The NIH All of Us Pro- gram, the Million Veteran Project (MVP) sponsored by US Department of Veterans Affairs (VA), and the UK Biobank are three prime examples of recent massive datasets. These datasets require terabytes of storage on sample sizes ranging from 105 to 106 and above subjects. The datasets are also dynamic, growing over time in size and complexity. In addition, the datasets are heterogeneous; for example, the UK Biobank offers ge- nomic data, electronic health record (EHR) data, and imaging data on the same study individuals. Finally, as with most real-world data, the data are fraught with missingness and inaccuracy. We propose attacking the issues of parameter estimation and model selection raised by such massive datasets. We will be guided by princi- ples of parsimony and high-dimensional optimization. Most of the speci?c applications we have in mind involve imaging and genomics, particularly genomewide association discovery. Fortunately, most of the tools and soft- ware we construct will be more generically useful. Our successful algo- rithms will be coded in the modern scienti?c programming language Julia and posted on publicly available websites. We will focus on constrained and sparse regression, EM and MM algorithms for optimization, variance components models, bootstrapping of linear mixed models, a copula-like model for correlated data, and sensitivity analysis in epidemic models. These are all subjects of paramount importance in modern genomics, bio- statistics and data mining.