Sheng Zhong, Ph.D. - US grants

NSF 0524139 / 0524030

Collaborative Research: Incentive-Compatible Protocols

Rebecca N. Wright, Stevens Institute and Sheng Zhong, State University of New York at Buffalo

The rapid expansion of the Internet has changed the way we communicate, with ever-increasing aspects of our daily life involving computation and data communication. However, many communications protocol designs assume that all participants will follow the specified protocols, and do not maintain their desirable properties if they don't. One approach to this problem is to design for a world in which some parties may maliciously deviate from their protocols in arbitrary ways. However, the resulting designs are typically expensive or impossible to achieve, and in many cases overestimate the kinds of misbehavior one is likely to encounter. Weaker security goals are more achievable, but often underestimate the willingness of participants to cheat.

This project studies an intermediate notion of security, known as incentive compatibility. In many practical settings, it is reasonable to assume that users will tend to act in their own best interests. Therefore, if one can design and use protocols that are incentive-compatible, in that following the protocol produces outcomes for participants at least as good as deviating from the protocol, then one can in many settings avoid misbehavior. Incentive compatibility, which borrows ideas from economics and game theory, is a more finely tunable notion of security than traditional security models, allowing different kinds of participants motivated by different incentives. Such solutions can provide better trust and assurance than protocols secure only against passive adversaries, at a potentially lower cost than protocols secure against an arbitrarily malicious adversary. This allows incentive-compatible protocols to provide the "right" level of security, at the "right" cost.

Four areas studied are: incentive-compatible ad hoc networks, incentive-compatible data mining, (both of which can significantly benefit from introducing incentive-compatible solutions), testing tools for incentive compatibility, and foundations of incentive compatibility.

(This award is funded through the American Recovery and Reinvestment Act of 2009: Public Law 111-5).

This is a CAREER award to support the research of Dr. Sheng Zhong, in the Department of bioengineering at University of Illinois at Urbana-Champaign. Dr. Zhong is a fourth-year, tenure-track Assistant Professor. Understanding cellular processes, such as development and differentiation, requires understanding how cells control the rate of protein production, especially gene transcription. A gene regulatory network (GRN) is a collection of DNA segments in a genome, which interact with each other and with other substances in the cell, thereby governing the rates at which genes in a network are transcribed into messenger RNA. Most computational tools for identification of eukaryote GRNs were developed for the single cell eukaryote, Saccharomyces cerevisiae (baker's yeast). The complexity of GRNs in multicellular organisms and especially in vertebrates is several magnitudes greater than that of the yeast GRNs. The greater complexity arises from how genes are controlled by complex patterns of DNA regulatory elements, called cis-regulatory modules; how larger genome size provides a greater template for cis-regulatory modules to evolve; and the stochastic process of transcription, which is activated by the interaction of regulatory proteins and cis-regulatory modules. A new framework for the identification of GRNs in multicellular organisms is being developed by Dr. Zhong. Quantitative evolution models and joint models of gene expression and protein-DNA interaction data are being developed in this project. The project is advancing the state of the art of systems biology through developing a theoretical framework of eukaryote GRN evolution, a variety of GRN identification and analysis methods, prototype systems for analysis of genomic data, and through discovery of engineering principles in cell biology. Databases and tools produced under this project will be accessible via the PI?s website at http://bioinformatics.bioen.uiuc.edu/

As a part of his CAREER plan, Dr. Zhong is training a new generation of interdisciplinary researchers by engaging undergraduate students in research; developing new courses and participating in outreach activities with Illinois middle school and high school students, including a bioinformatics camp for middle school girls, as one of the Girls Adventures in Mathematics, Engineering, and Science (G.A.M.E.S) camp program.

Recent advances in computing hardware and wireless technology have resulted in widespread popularity of applications and systems involving highly distributed wireless networks. In such a network, each node may belong to an independent organization or individual that has its own interests, and so may not always want to behave cooperatively. However, if these nodes do not cooperate, the performance of the network may degrade, or the entire network may become dysfunctional. Consequently, we need to design economic mechanisms to stimulate these nodes to cooperate.

In this project, we study the design of enforceable economic mechanisms. We say an economic mechanism is enforceable if it provides sufficiently strong incentives
for cooperative behavior and security protection against cheating and has been systematically evaluated in experiments. We emphasize the great importance to make economic mechanisms enforceable because if an economic mechanism is not enforceable, it cannot be effectively used in practice to simulate cooperation.

Our research plan is to focus on two specific problems, namely the Routing and Packet Forwarding Problem and the FDMA Channel Assignment Problem, because they are simple but fundamental. Besides these two problems, it is also planned to study experimental methods for systematically evaluating the incentive compatibility and computation and communication efficiency of economic mechanisms in wireless networks, as well as a few other selected problems of non-cooperation in wireless networks. In addition, the plan includes education activities to disseminate cutting-edge knowledge on cooperation in wireless networks to undergraduate and graduate students.

Humans and other organisms are enormously genetically variable, even for traits like lifespan and fertility. During the past decade, genes affecting longevity and fertility have been discovered by mutating those genes in laboratory animals and observing the resulting effects. For example, mutations in genes controlling insulin signaling can increase the longevity of worms, flies, and mice by 100% or more. These results beg the question of whether the same genes are responsible for natural variation among individuals, populations, and species in longevity and fertility. This project addresses this question by discovering the genetic basis of life history differences in artificially-selected and in natural populations of an insect, Drosophila melanogaster, for which powerful genetic techniques are available. Expression profiling, genetic mapping, and new statistical techniques will be deployed to identify the genes contributing to longevity and fertility differences between populations.

Genes causing variation in longevity and fertility have significant impact on human health and well being. Most of the proposed candidate genes are shared between insects and mammals (including humans), and knowing which genes cause variation in non-inbred animals has important implications for biomedicine. Other significant effects will include training of undergraduate and graduate researchers in the laboratories of all three investigators, and at two ?mini-symposia? that will include all participants, and development of new analytical tools.

DESCRIPTION (provided by applicant): PROJECT SUMMARY / ABSTRACT: It is well recognized that genetic aberrations, such as chromosomal translocations, are complemented by secondary aberrations (possible epigenetic) that give rise to leukemic clones. However, complementary epigenetic event(s) and their roles in development of common subtypes of childhood leukemias are not well characterized. Our goals are to understand the scope of genome-wide epigenetic alterations in childhood leukemia, and then develop epigenetic screening markers to test archived neonatal blood spots (Guthrie cards) for early detection of leukemia. We have successfully performed preliminary experiments using GoldenGate Methylation Cancer Panel I microarrays (Illumina, Inc), using leukemic bone marrows, and corresponding archived Guthrie cards of childhood leukemias patients, to test the contribution of a large panel of cancer- related genes in the development of leukemia. We have also performed experiments with leukemia cell lines to test the gene expression response to a demethylating agent. This technique involved pharmacologic inhibition of epigenetic modifications, coupled with gene expression microarrays (Affymetrix Exon 1.0), and has yielded a number of candidate genes for which gene promoter methylation may play a functional role in leukemia. Our current focus is to focus on a panel of preliminary candidate genes using validation assays in primary leukemia samples, and then focusing further in Guthrie cards for our chosen targets. In the current application we will use methylation-specific PCR and bisulfite sequencing to reduce the size, and to focus our current large panel of 90 gene targets. We will choose the 30 which are methylated in the highest fraction of TEL-AML1 leukemia and which are not methylated in normal CD34 cells and normal Guthrie card DNA. We will then test this panel of epigenetic markers in Guthrie cards from children who grew up to get TEL-AML1 leuekmia and compare to a group of 500 children who did not, to assess predictive value. We focus exclusively on the most common cytogenetic subltype, those with TEL-AML1 translocations who are 22% of all ALL. This research aims ultimately at the characterization of early epigenetic mutations and imprinting abnormalities in the development of leukemia, and the use these markers for the early detection of pediatric leukemia. PUBLIC HEALTH RELEVANCE: PROJECT NARRATIVE: With this project we are exploring the existence of early markers of leukemia at birth. This will impact public health by increasing our knowledge in the mechanisms of the etiology of childhood leukemia and providing markers for the early detection of, or predisposition to, leukemia at birth.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).

A recently developed technology, called opportunistic coding, can greatly improve the e±ciency of wireless networks. However, the problems of cooperation and security in opportunistic-coding-based (OCB) wireless networks have not received su±cient attention. Here cooperation problems
are the problems introduced by the existence of sel¯sh nodes in the wireless networks, while security problems are the problems introduced by the existence of adversarial nodes. This work is to design and implement solutions that can provide cooperation and security guarantees for OCB wireless networks. The research work can be divided into studied of three problem areas, namely the cooperation problems, the security problems, and the interplays of
cooperation and security. The work also includes the dissemination of related knowledge to students at various levels.

Intellectual Merit: The results of this work will signi¯cantly improve OCB wireless networks in terms of cooperation and security. Furthermore, these results will also have broader theoretical interests, because a part of the techniques to be developed will be applicable to economic incentives
problems in other settings as well.

Broader Impacts: First, the developed technical solutions will bene¯t the society because they make it possible to widely deploy OCB wireless networks to environments with sel¯sh or adversarial nodes. Second, the educational component of this work will build capacity in cooperation and security at various levels.

The University of Illinois at Urbana-Champaign is awarded a grant to advance the understanding of the evolution of complex traits, such as the development of multi-cellular body plans or an organism's social behavior. The PI will use the analysis of brain plasticity and social behavior as test-bed questions to build and test quantitative genomic models for the evolution of complex traits. Brain plasticity refers to the brain's ability to reorganize itself by forming new neural connections as a result of experience. The project will test the hypothesis that genetic pathways that mediate brain plasticity are enriched in conserved gene regulatory modules in brain tissues across species, by developing novel analytical models and computational tools for modeling the evolution of gene regulatory modules. The first model is an evolutionary model for genetic modules, which can be applied to identify conserved as well as species-specific genetic modules using and gene expression data in multiple species and phylogenetic distances. The second model can be applied to analyze the transcription networks implemented in related species with a conserved phenotype. The regulatory relationships between an orthologous set of TFs and target genes will be simultaneously modeled and identified in all the species under consideration. Both sequence data and gene expression data, when available, will be modeled. This project will deliver two software tools for integrated comparative analysis of genome and transcriptome data. by These software tools will be hosted on a dedicated server (http://sysbio.bioen.uiuc.edu/grn.htm) and made available to the entire research community through user-friendly web applications. The project will include the participation of undergraduate, graduate, and postdoctoral students, and the software will be used in a hands-on course taught by the investigator at a bioinformatics camp for young women.

A major goal of biology is to identify the processes that shape the evolution of form and structure. To pursue this important goal, this project will take an innovative, comparative approach to investigate how evolutionary changes during limb development (the process by which the limb forms) contribute to the generation of the divergent limbs of three mammals: bats, opossums, and mice. The mammal limb is an ideal system with which to pursue this goal because the way that a mammal feeds, moves and behaves is dependent upon the form of its limbs. As such, from the wings of bats to the flippers of whales to the hooves of horses, the diversification of the mammal limb has been crucial to the ecological and evolutionary success of the group. Through use of traditional and next generation genetic analyses, this project will identify the specific changes in gene expression (e.g., when genes are turned on and off, and at what levels) that initiate and drive the divergence of limb form among mammalian species. This knowledge will vertically advance our understanding of how the process of development constrains and facilitates the evolution of certain limb forms, and thereby impacts mammalian evolution. Furthermore, as humans are mammals, this project will directly advance our understanding of the processes that have shaped our own evolutionary history.

DESCRIPTION (provided by applicant): RNA-RNA interactions are important components of gene regulatory networks. There is yet no technology to map the entire RNA-RNA interactome in cells or tissues. As a result, although personal genomes are being sequenced, our capabilities to interpret genomic functions and to predict phenotypic variation from genomic sequence remain limited. We desperately need high-throughput technologies to map the molecular networks in any person or tissue. Here, we propose to develop an extremely high-throughput technology to map RNA-RNA interactomes in vivo. The central idea is to convert molecular interactions into DNA sequences and then read out the interactions by DNA sequencing. The proposed RNA Hi-C technology is capable of mapping all protein- assisted RNA-RNA interactions in one assay. This new technology is expected to be generally applicable to analyze any cell types and tissues. We will use cell fate decisions in pre-implantation development as a driven biological question. The new technology will be applied to test two competing hypotheses. We expect the results to clarify a physical principle behind the earliest cell fate decision in mammals, and therefore offer unprecedented information regarding infertility, miscarriage, and birth defects.

? DESCRIPTION (provided by applicant): The genome of each individual harbors millions of nucleotide variants, and a major challenge is to understand how these variants contribute to phenotypic variations in the population. We propose a combined computational and experimental framework for identifying non-coding variants that affect cellular and physiological traits, with the goal to establish computational models that can predict the probability of exhibiting a physiological trait from the sequences of non-coding genomic regions. This framework involves iterative refinement of model assumptions and parameters with experimentation. To develop the framework and validate the predictive models, we will focus on the disease Age-related Macular Degeneration (AMD), the leading cause of blindness among the elderly in the country. Previous studies have identified a number of sequence variants strongly associated with AMD. We will develop computational models to predict (or narrow down) the set of non-coding sequence variants that contribute to the disease phenotype. As experimental assessment, we will perform genome editing in patient-derived induced pluripotent stem cells (iPSC) to test the consequence of removing or introducing such sequence variants on molecular and cellular phenotypes in cell culture and in rodent models. While the proposed method is developed for AMD, the general approach is expected to apply to other genetic diseases.

? DESCRIPTION (provided by applicant): Recent technological developments have significantly advanced our understanding of the three-dimensional organization of the nucleus. It has also become increasingly clear that the 3D nucleome plays an important role in regulating gene expression. Large cohorts of data are being generated to investigate the impact of the temporal changes of nuclear organization (the 4D nucleome) to normal development and disease processes. The primary goal of this proposal is the creation of an effective organizational hub and web portal for the 4D nucleome. We also propose to develop an effective coordination structure for the 4D nucleome activities. First, we will develop and organize effective networking methods and consortium meetings for establishing protocols and standards. We will develop a framework for coordination of the funded 4D nucleome projects. We will organize annual consortium-wide grantees meeting. We will utilize the professional organizers who have previously worked with the PIs and the premium conference venues at Atkinson Hall in UCSD (http://www.calit2.net). Two major activities will be organized during the annual meeting. Second, we will develop an adaptable, scalable, and user-friendly 4DN web portal. We will build the 4DN Network community web portal and a Virtual Resource Repository. This portal will be an integrated, versatile, and inter-operable data management, retrieval, analysis, and visualization system. We will leverage the ENCODE Comparative Browser (http://encode.cepbrowser.org) for developing the 4D network portal. In addition, the portal will incorporate data generated from outside the consortium, publish E-manuals for the consortium-agreed protocols, and provide the most up-to-date information on the data clearance. Third, we will coordinate and manage the 4DN Network Opportunity Pool. The opportunity pool of funds will be distributed and managed with 100% transparency. Competitive distribution with the funds will be coordinated by a professional manager with extensive experience in managing Center and Training Grants. Fourth, we will develop a comprehensive strategy for training and outreach. The proposed unit will aim to regularly train and update 4DN research network members and collaborators in several key areas including: 1) new technologies; 2) data/sample collection protocols; 3) data analysis methods; 4) data submission protocols; 5) data retrieval methods; 6) data QA/QC; 7) data exchange standards and data ontologies; 8) the contents of different 4DN and public databases; 9) data privacy and 10) data dissemination.

Extremely high-throughput mapping of protein, RNA, and chromatin interactions in health and disease Abstract This Catalyst project aims to removing a major bottleneck in understanding diabetes and its complications, by developing the technologies to map diverse molecular interactions in the disease-relevant cells at the genomic scale. These proposed technologies, collectively called PRACI (Protein, RNA, and chromatin interactions), will enable a typical research lab to map genome-wide protein-protein, RNA-protein, RNA-RNA, and RNA- DNA/chromatin interaction networks from a given cell type within 1 months? time. PRACI enables typical labs to compare molecular interaction networks between health and disease states. Without PRACI, genome-wide mapping of even a single type of interactions from a disease-relevant cell type remains a formidable task. I will systematically map molecular interactome changes related to diabetes related vascular complications, using hyperglycemia and chronic inflammation-induced irreversible alterations vascular endothelial cells as a testbed system. I anticipate to reveal which components of the multiscale molecular networks are responsible for the sustained dysregulation of gene expression in dysfunctional endothelial cells. Such information will lead to new perspectives to diabetic wound healing, given the established roles of endothelial dysfunction to diabetic wounds and the relative accessibility of vasculature. I anticipate that these technologies and their enabled discoveries will contribute to and inspire transformative changes in the study of Diabetes, Endocrinology, and Metabolic Diseases.

Spatial in situ mapping of RNA-chromatin interactions in human vasculature Abstract A breakthrough of this project is to simultaneously improve the state-of-art spatial resolution, dynamic range of RNA quantification, and applicable dimensions of human tissue by one to several orders of magnitude. This breakthrough will deliver to users extremely high-resolution, high-density, high-fidelity and reproducible spatial mapping slides (HiFi slides) for spatial transcriptome analysis or spatial mapping of modified- or interacting- nucleic acids. We will leverage HiFi slides to provide spatial transcriptome and spatial RNA-chromatin interactome mapping tools in unprecedented resolution. This project will generate HiFi spatial transcriptome, HiFi spatial RNA-crhomatin interactome, and high-resolution protein spatial profiles of 96 proteins from mesenteric arteries of veins, and the microvasculature from pancreas, liver and brain. These datasets will form the initial human EC biomolecular map (HuEC-MAP). These data will offer unprecedented insights to the molecular nature of the tissue-specific morphologies and functions of the blood vessel endothelium, which are critical to the development and functions of the associated tissues.

Revealing protein-protein interactions and RNA-protein interactions at genome-scale in two weeks Abstract Our ability to interpret the human genome is limited by our knowledge of the interaction networks of the products of the genome sequence, including RNAs and proteins. If we had a complete reference map of all DNA-DNA, protein-DNA, RNA-DNA, RNA-RNA, RNA-protein, and protein-protein interactions, we would have a completely new approach to reading the book of the human genome. Recent technology breakthroughs, including those led by NIH Common Fund programs, enabled genome-wide mapping of DNA-DNA, protein-DNA, RNA-DNA, and RNA-RNA interactions en masse. However, genome-scale mapping of RNA-protein interactions (RPI) and protein-protein interactions (PPI) remain laborious and resource-intensive. In this project, we propose two extremely high-throughput genomic-based technologies for mapping human RPI and PPI networks at genome scale. We will develop bioinformatics tools to analyze the data with statistical rigor. Specifically, we propose an ?all-vs-all? approach to map PPI and RPI networks. These genomics technologies and their coupled genomic informatics tools will generate reference maps of human PPI and RPI networks. In the long term, such maps will greatly facilitate the interpretation of the functions of the human genome. In Aim 1, we will develop an extremely high-throughput technology called ?PPI-seq? to map the PPI network at genome-scale. PPI-seq is expected to be capable of generating a reference map of the human PPI network within a single lab in 2 weeks' time, with an expected yield of ~50,000 high-confidence pairwise PPIs. In Aim 2, we will develop an extremely high- throughput technology called ?RPI-seq? to map the RPI network at genome-scale. RPI-seq is expected to be capable of generating a reference map of the human RPI network within single lab in 2 weeks' time. RPI-seq will simultaneous reveal RNA binding proteins (RBP) and the RNAs bound by each RBP.

The second phase of NIH Common Fund 4D Nucleome Network Organizational Hub Abstract The second phase of the 4DN program requires an efficient organizational center to synergize the activities of all the funded teams and integrate the research products. Accordingly, the 4DN Organizational Hub is proposed to: (1) coordinate and integrate the efforts of all the funded projects, (2) build an efficient consortium infrastructure to serve as a center for the collaborative efforts, and (3) provide the 4DN web portal as a central resource gateway to access all the 4DN-Network generated tools, policies, guidelines, protocols, reagents, cell lines, and as a central hub for the outreach activities. Our major deliverables are: (1) an organizational structure composed of a steering committee and problem-solving working groups with a clear report chain to enable effective communications and decision-making processes, establish co-organizers and schedule, clarify action items, and ensure execution of the action items; (2) 4DN Web Portal (https://www.4dnucleome.org/), which is the always up-to-date community-wide resource and point of access for all data, protocols, reagents, resources, and methods; (3) 4DN internal wiki, the central organized resource for all teams, centers, and working groups to document internal progresses. (4) 4DN annual meetings, including the kick-off meeting in winter 2020 and the subsequent annual meetings and 4DN-ASCB (American Society of Cell Biology) satellite meetings; (5) 4DN outreach workshops at Keystone symposia and American Society of Human Genetics (ASHG) meetings, and on 4DN YouTube channel.