SURVIVAL PREDICTION FOR BRAIN TUMOR PATIENTS USING GENE EXPRESSION DATA

Brain tumor is one of the most aggressive types of cancer in humans, with an estimated median survival time of 12 months and only 4% of the patients surviving more than 5 years after disease diagnosis. Until recently, brain tumor prognosis has been based only on clinical information such as tumor grade and patient age, but there are reports indicating that molecular profiling of gliomas can reveal subgroups of patients with distinct survival rates. We hypothesize that coupling molecular profiling of brain tumors with clinical information might improve predictions of patient survival time and, consequently, better guide future treatment decisions. In order to evaluate this hypothesis, the general goal of this research is to build models for survival prediction of glioma patients using DNA molecular profiles (U133 Affymetrix gene expression microarrays) along with clinical information. First, a predictive Random Forest model is built for binary outcomes (i.e. short vs. long-term survival) and a small subset of genes whose expression values can be used to predict survival time is selected. Following, a new statistical methodology is developed for predicting time-to-death outcomes using Bayesian ensemble trees. Due to a large heterogeneity observed within prognostic classes obtained by the Random Forest model, prediction can be improved by relating time-to-death with gene expression profile directly. We propose a Bayesian ensemble model for survival prediction which is appropriate for high-dimensional data such as gene expression data. Our approach is based on the ensemble "sum-of-trees" model which is flexible to incorporate additive and interaction effects between genes. We specify a fully Bayesian hierarchical approach and illustrate our methodology for the CPH, Weibull, and AFT survival models. We overcome the lack of conjugacy using a latent variable formulation to model the covariate effects which decreases computation time for model fitting. Also, our proposed models provides a model-free way to select important predictive prognostic markers based on controlling false discovery rates. We compare the performance of our methods with baseline reference survival methods and apply our methodology to an unpublished data set of brain tumor survival times and gene expression data, selecting genes potentially related to the development of the disease under study. A closing discussion compares results obtained by Random Forest and Bayesian ensemble methods under the biological/clinical perspectives and highlights the statistical advantages and disadvantages of the new methodology in the context of DNA microarray data analysis.

A PROCESS FOR ACHIEVING COMPARABLE DATA FROM HETEROGENEOUS DATABASES

The current state of health and biomedicine includes an enormity of heterogeneous data ‘silos’, collected for different purposes and represented differently, that are presently impossible to share or analyze in toto. The greatest challenge for large-scale and meaningful analyses of health-related data is to achieve a uniform data representation for data extracted from heterogeneous source representations. Based upon an analysis and categorization of heterogeneities, a process for achieving comparable data content by using a uniform terminological representation is developed. This process addresses the types of representational heterogeneities that commonly arise in healthcare data integration problems. Specifically, this process uses a reference terminology, and associated "maps" to transform heterogeneous data to a standard representation for comparability and secondary use. The capture of quality and precision of the “maps” between local terms and reference terminology concepts enhances the meaning of the aggregated data, empowering end users with better-informed queries for subsequent analyses. A data integration case study in the domain of pediatric asthma illustrates the development and use of a reference terminology for creating comparable data from heterogeneous source representations. The contribution of this research is a generalized process for the integration of data from heterogeneous source representations, and this process can be applied and extended to other problems where heterogeneous data needs to be merged.

Three linked enzyme loci in fishes: implications in the evolution of vertebrate chromosomes.

A three-point linkage group comprised of loci coding for adenosine deaminase (ADA), glucose-6-phosphate dehydrogenase (G6PDH), and 6-phospho-gluconate dehydrogenase (6PGD) is described in fish of the genus Xiphophorus (Poeciliidae). The alleles at loci in this group were shown to assort independently from the alleles at three other loci--isocitrate dehydrogenase 1 and 2, and glyceraldehyde-3-phosphate dehydrogenase 1. Alleles at the latter three loci also assort independently from each other. Data were obtained by observing the segregation of electrophoretically variant alleles in reciprocal backcross hybrids derived from crosses between either X. helleri guentheri or X. h. strigatus and X. maculatus. The linkage component of chi2 was significant (less than 0.01) in all crosses, indicating that the linkage group is conserved in all populations of both species of Xiphophorus examined. While data from X. h. guentheri backcrosses indicate the linkage relationship ADA--6%--G6PDH--24%--6PGD, and ADA--29%--6PGD (30% when corrected for double crossovers), data from backcrosses involving strigatus, while supporting the same gene order, yielded significantly different recombination frequencies. The likelihood of the difference being due to an inversion could not be separated from the possibility of a sex effect on recombination in the present data. The linkage of 6PGD and G6PDH has been shown to exist in species of at least three classes of vertebrates, indicating the possibility of evolutionary conservation of this linkage.

THE INTERACTION BETWEEN INTERNAL AND EXTERNAL INFORMATION ON RELATIONAL DATA SEARCH

People often use tools to search for information. In order to improve the quality of an information search, it is important to understand how internal information, which is stored in user’s mind, and external information, represented by the interface of tools interact with each other. How information is distributed between internal and external representations significantly affects information search performance. However, few studies have examined the relationship between types of interface and types of search task in the context of information search. For a distributed information search task, how data are distributed, represented, and formatted significantly affects the user search performance in terms of response time and accuracy. Guided by UFuRT (User, Function, Representation, Task), a human-centered process, I propose a search model, task taxonomy. The model defines its relationship with other existing information models. The taxonomy clarifies the legitimate operations for each type of search task of relation data. Based on the model and taxonomy, I have also developed prototypes of interface for the search tasks of relational data. These prototypes were used for experiments. The experiments described in this study are of a within-subject design with a sample of 24 participants recruited from the graduate schools located in the Texas Medical Center. Participants performed one-dimensional nominal search tasks over nominal, ordinal, and ratio displays, and searched one-dimensional nominal, ordinal, interval, and ratio tasks over table and graph displays. Participants also performed the same task and display combination for twodimensional searches. Distributed cognition theory has been adopted as a theoretical framework for analyzing and predicting the search performance of relational data. It has been shown that the representation dimensions and data scales, as well as the search task types, are main factors in determining search efficiency and effectiveness. In particular, the more external representations used, the better search task performance, and the results suggest the ideal search performance occurs when the question type and corresponding data scale representation match. The implications of the study lie in contributing to the effective design of search interface for relational data, especially laboratory results, which are often used in healthcare activities.

Upper ocean vertical supply: A neglected primary factor controlling the distribution of neodymium concentrations of open ocean surface waters?

Neodymium (Nd) isotopes are an important geochemical tool to trace the present and past water mass mixing as well as continental inputs. The distribution of Nd concentrations in open ocean surface waters (0�100 m) is generally assumed to be controlled by lateral mixing of Nd from coastal surface currents and by removal through reversible particle scavenging. However, using 228Ra activity as an indicator of coastal water mass influence, surface water Nd concentration data available on key oceanic transects as a whole do not support the above scenario. From a global compilation of available data, we find that more stratified regions are generally associated with low surface Nd concentrations. This implies that upper ocean vertical supply may be an as yet neglected primary factor in determining the basin-scale variations of surface water Nd concentrations. Similar to the mechanism of nutrients supply, it is likely that stratification inhibits vertical supply of Nd from the subsurface thermocline waters and thus the magnitude of Nd flux to the surface layer. Consistently, the estimated required input flux of Nd to the surface layer to maintain the observed concentrations could be nearly two orders of magnitudes larger than riverine/dust flux, and also larger than the model-based estimation on shelf-derived coastal flux. In addition, preliminary results from modeling experiments reveal that the input from shallow boundary sources, riverine input, and release from dust are actually not the primary factors controlling Nd concentrations most notably in the Pacific and Southern Ocean surface waters.

NETWORK TOPOLOGY IN HUMAN PROTEIN INTERACTION DATA PREDICTS FUNCTIONAL ASSOCIATION

High-throughput assays, such as yeast two-hybrid system, have generated a huge amount of protein-protein interaction (PPI) data in the past decade. This tremendously increases the need for developing reliable methods to systematically and automatically suggest protein functions and relationships between them. With the available PPI data, it is now possible to study the functions and relationships in the context of a large-scale network. To data, several network-based schemes have been provided to effectively annotate protein functions on a large scale. However, due to those inherent noises in high-throughput data generation, new methods and algorithms should be developed to increase the reliability of functional annotations. Previous work in a yeast PPI network (Samanta and Liang, 2003) has shown that the local connection topology, particularly for two proteins sharing an unusually large number of neighbors, can predict functional associations between proteins, and hence suggest their functions. One advantage of the work is that their algorithm is not sensitive to noises (false positives) in high-throughput PPI data. In this study, we improved their prediction scheme by developing a new algorithm and new methods which we applied on a human PPI network to make a genome-wide functional inference. We used the new algorithm to measure and reduce the influence of hub proteins on detecting functionally associated proteins. We used the annotations of the Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) as independent and unbiased benchmarks to evaluate our algorithms and methods within the human PPI network. We showed that, compared with the previous work from Samanta and Liang, our algorithm and methods developed in this study improved the overall quality of functional inferences for human proteins. By applying the algorithms to the human PPI network, we obtained 4,233 significant functional associations among 1,754 proteins. Further comparisons of their KEGG and GO annotations allowed us to assign 466 KEGG pathway annotations to 274 proteins and 123 GO annotations to 114 proteins with estimated false discovery rates of <21% for KEGG and <30% for GO. We clustered 1,729 proteins by their functional associations and made pathway analysis to identify several subclusters that are highly enriched in certain signaling pathways. Particularly, we performed a detailed analysis on a subcluster enriched in the transforming growth factor β signaling pathway (P<10-50) which is important in cell proliferation and tumorigenesis. Analysis of another four subclusters also suggested potential new players in six signaling pathways worthy of further experimental investigations. Our study gives clear insight into the common neighbor-based prediction scheme and provides a reliable method for large-scale functional annotations in this post-genomic era.

Analysis of Band 4.1B in Integrin-Mediated Cell Adhesion and Signaling

Band 4.1B is a cytoskeletal adaptor protein that regulates various cellular behavior; however, the mechanisms by which Band 4.1B contributes to intracellular signaling are unclear. This project addresses in vivo and in vitro functions for Band 4.1B in integrin-mediated cell adhesion and signaling. Band 4.1B has been shown to bind to β8 integrin, although cooperative functions of these two proteins have not been determined. Here, functional links between β8 integrin and Band 4.1B were investigated using gene knockout strategies. Ablation of β8 integrin and Band 4.1B genes resulted in impaired cardiac morphogenesis, leading to embryonic lethality by E11.5. These embryos displayed malformation of the outflow tract that was likely linked to abnormal regulation of cardiac neural crest migration. These data indicate the importance of cooperative signaling between β8 integrin and Band 4.1B in cardiac development. The involvement of Band 4.1B in integrin-mediated cell adhesion and signaling was further demonstrated by studying its functional roles in vitro. Band 4.1B is highly expressed in the brain, but its signaling in astrocytes is not understood. Here, Band 4.1B was shown to promote cell spreading likely by interacting with β1 integrin via its band 4.1, ezrin, radixin, and moesin (FERM) domain in cell adhesions. In astrocytes, both Band 4.1B and β1 integrin were expressed in cell-ECM contact sites during early cell spreading. Exogenous expression of Band 4.1B, especially its FERM domain, enhanced cell spreading on fibronectin, an ECM ligand for β1 integrin. However, the increased cell spreading was prohibited by blocking β1 integrin. These findings suggest that Band 4.1B is crucial for early adhesion assembly and/or signaling that are mediated by β1 integrin. Collectively, this study was the first to establish Band 4.1B as a modulator of integrin-mediated adhesion and signaling.

CELL POLARITY REGULATES ORGAN GROWTH THROUGH THE HIPPO PATHWAY

Defects in apical-basal cell polarity and abnormal expression of cell polarity determinants are linked to human cancer. Loss of polarity is highly correlated with malignancy. In Drosophila, perturbation of apical-basal polarity, including overexpressing the apical determinant Crumbs, can lead to uncontrolled tissue growth. Cells mutant for the basolateral determinant scribble overproliferate and can form neoplastic tumors. Interestingly, scribble mutant clones that arise in wild-type tissues are eliminated and therefore do not manifest their tumorigenic potential. However, the mechanisms by which cell polarity coordinates with growth control pathways in developing organs to achieve appropriate organ size remain obscure. To investigate the function of apical determinants in growth regulation, I investigated the mechanism by which the apical determinant Crumbs affects growth in Drosophila imaginal discs. I found that crumbs gain and loss of function cause overgrowth and induction of Hippo target genes. In addition, Crumbs is required for the proper localization of Expanded, an upstream component of the Hippo pathway. Furthermore, we uncoupled the cell polarity and growth control function of Crb through structure-functional analysis. Taken together, our data identify a role of Crb in growth regulation specifically through modulation of the Hippo pathway. To further explore the role of polarity in growth control, I investigated how cells mutant for basolateral determinants are eliminated by using patches of cells mutant for scribble (scribble mutant clones) as a model system. We found that competitive cell-cell interactions eliminate tumorigenic scribble cells by modulation of the Hippo pathway. The regulation of Hippo signaling is required and sufficient to restrain the tumorous growth of scribble mutant cells. Artificially increasing the relative fitness of scribble mutant cells unleashes their tumorigenic potential. Therefore, we have identified a novel tumor-suppression mechanism that depends on signaling between normal and tumorigenic cells. These data identify evasion of cell competition as a critical step toward malignancy and illustrate a role for wild-type tissue in eliminating abnormal cells and preventing the formation of tumors.

The effect of idursulfase on growth in patients with Hunter syndrome: data from the Hunter Outcome Survey (HOS)

Hunter syndrome (mucopolysaccharidosis type II) is a rare and life-limiting multisystemic disorder with an X-linked recessive pattern of inheritance. Short stature is a prominent feature of this condition. This analysis aimed to investigate the effects of enzyme replacement therapy with idursulfase on growth in patients enrolled in HOS - the Hunter Outcome Survey which is a multinational observational database. As of Jan 2012, height data before treatment were available for 567 of 740 males followed prospectively after HOS entry. Cross-sectional analysis showed that short stature became apparent after approximately 8 years of age; before this, height remained within the normal range. Age-corrected standardized height scores (z-scores) before and after treatment were assessed using piecewise regression model analysis in 133 patients (8-15 years of age at treatment start; data available on ≥ 1 occasion within +/-24 months of treatment start; growth hormone-treated patients excluded). Results showed that the slope after treatment (slope=-0.005) was significantly improved compared with before treatment (slope=-0.043) (difference=0.038, p=0.004). Analysis of covariates (age at treatment start, cognitive involvement, presence of puberty at the start of ERT, mutation type, functional classification), showed a significant influence on growth of mutation type (height deficit in terms of z-scores most pronounced in patients with deletions/large rearrangements/nonsense mutations, p<0.0001) and age (most pronounced in the 12-15-year group, p<0.0001). Cognitive involvement, pubertal status at the start of ERT and functional classification were not related to the growth deficit or response to treatment. In conclusion, the data showed an improvement in growth rate in patients with Hunter syndrome following idursulfase treatment.

EVALUATION OF INTENSITY MODULATED RADIATION THERAPY (IMRT) DELIVERY ERROR DUE TO IMRT TREATMENT PLAN COMPLEXITY AND IMPROPERLY MATCHED DOSIMETRY DATA

Intensity modulated radiation therapy (IMRT) is a technique that delivers a highly conformal dose distribution to a target volume while attempting to maximally spare the surrounding normal tissues. IMRT is a common treatment modality used for treating head and neck (H&N) cancers, and the presence of many critical structures in this region requires accurate treatment delivery. The Radiological Physics Center (RPC) acts as both a remote and on-site quality assurance agency that credentials institutions participating in clinical trials. To date, about 30% of all IMRT participants have failed the RPC’s remote audit using the IMRT H&N phantom. The purpose of this project is to evaluate possible causes of H&N IMRT delivery errors observed by the RPC, specifically IMRT treatment plan complexity and the use of improper dosimetry data from machines that were thought to be matched but in reality were not. Eight H&N IMRT plans with a range of complexity defined by total MU (1460-3466), number of segments (54-225), and modulation complexity scores (MCS) (0.181-0.609) were created in Pinnacle v.8m. These plans were delivered to the RPC’s H&N phantom on a single Varian Clinac. One of the IMRT plans (1851 MU, 88 segments, and MCS=0.469) was equivalent to the median H&N plan from 130 previous RPC H&N phantom irradiations. This average IMRT plan was also delivered on four matched Varian Clinac machines and the dose distribution calculated using a different 6MV beam model. Radiochromic film and TLD within the phantom were used to analyze the dose profiles and absolute doses, respectively. The measured and calculated were compared to evaluate the dosimetric accuracy. All deliveries met the RPC acceptance criteria of ±7% absolute dose difference and 4 mm distance-to-agreement (DTA). Additionally, gamma index analysis was performed for all deliveries using a ±7%/4mm and ±5%/3mm criteria. Increasing the treatment plan complexity by varying the MU, number of segments, or varying the MCS resulted in no clear trend toward an increase in dosimetric error determined by the absolute dose difference, DTA, or gamma index. Varying the delivery machines as well as the beam model (use of a Clinac 6EX 6MV beam model vs. Clinac 21EX 6MV model), also did not show any clear trend towards an increased dosimetric error using the same criteria indicated above.

EXPRESSION AND IMMUNOLOGICAL CHARACTERIZATION OF HERV-K TRANSMEMBRANE ENVELOPE PROTEIN IN HUMAN BREAST CANCER

The human endogenous retrovirus K (HERV-K) env gene encodes envelope protein comprising surface (SU) and transmembrane (TM) domains. Having shown the exclusive expression of SU in human breast cancer and the stimulation of SU-specific immune responses in patients with breast cancer, our research here confirmed and extended the data by investigating the expression of HERV-K TM envelope domain and the induction of specific immune responses against TM in breast cancer patients. We found HERV-K TM mRNA and protein expression only in human breast cancer cells but not in normal controls. The specific immune responses against TM domain were induced in mice determined by enzyme-linked immunosorbent assay (ELISA) and IFN-γ enzyme-linked immunosorbent spot (ELISPOT) assay. Furthermore, ELISA detected higher titers of anti-HERV-K TM Env IgG antibodies in sera of breast cancer patients. In addition, the magnitude of the anti-HERV TM B cell response was correlated with the disease stage. Peripheral blood mononuclear cells (PBMCs) before and after in vitro stimulation (IVS) with HERV-K TM from patients with breast cancer as well as healthy controls were tested for T cell responses against HERV-K TM domain by ELISPOT assay. Breast cancer patients (n=21) had stronger HERV-K TM-specific cellular responses than healthy controls (n=12) (P < 0.05). These findings suggest, for the first time, that HERV-K TM expression was enhanced in human breast cancer cells and was able to induce specific B cell and T cell immune responses in breast cancer patients. This study provides support for HERV-K TM as a promising source of antigen for anti-tumor immunotherapy, prevention, diagnosis, and prognosis.

HIV-related malignancies: A community-based study using a linkage of cancer registry and HIV registry data

Background. Increased incidence of cancer is documented in immunosuppressed transplant patients. Likewise, as survival increases for persons infected with the Human Immunodeficiency Virus (HIV), we expect their incidence of cancer to increase. The objective of this study was to examine the current gender specific spectrum of cancer in an HIV infected cohort (especially malignancies not currently associated with Acquired Immunodeficiency Syndrome (AIDS)) in relation to the general population.^ Methods. Cancer incidence data was collected for residents of Harris County, Texas who were diagnosed with a malignancy between 1975 and 1994. This data was linked to HIV/AIDS registry data to identify malignancies in an HIV infected cohort of 14,986 persons. A standardized incidence ratio (SIR) analysis was used to compare incidence of cancer in this cohort to that in the general population. Risk factors such as mode of HIV infection, age, race and gender, were evaluated for contribution to the development of cancer within the HIV cohort, using Cox regression techniques.^ Findings. Of those in the HIV infected cohort, 2289 persons (15%) were identified as having one or more malignancies. The linkage identified 29.5% of these malignancies (males 28.7% females 60.9%). HIV infected men and women had incidences of cancer that were 16.7 (16.1, 17.3) and 2.9 (2.3, 3.7) times that expected for the general population of Harris County, Texas, adjusting for age. Significant SIR's were observed for the AIDS-defining malignancies of Kaposi's sarcoma, non-Hodgkin's lymphoma, primary lymphoma of the brain and cancer of the cervix. Additionally, significant SIR's for non-melanotic skin cancer in males, 6.9 (4.8, 9.5) and colon cancer in females, 4.0 (1.1, 10.2) were detected. Among the HIV infected cohort, race/ethnicity of White (relative risk 2.4 with 95% confidence intervals 2.0, 2.8) or Spanish Surname, 2.2 (1.9, 2.7) and an infection route of male to male sex, with, 3.0 (1.9, 4.9) or without, 3.4 (2.1, 5.5) intravenous drug use, increased the risk of having a diagnosis of an incident cancer.^ Interpretation. There appears to be an increased risk of developing cancer if infected with the HIV. In addition to the malignancies routinely associated with HIV infection, there appears to be an increased risk of being diagnosed with non-melanotic skin cancer in males and colon cancer in females. ^

Volcanic forcing for climate modeling: a new microphysics-based data set covering years 1600–present

As the understanding and representation of the impacts of volcanic eruptions on climate have improved in the last decades, uncertainties in the stratospheric aerosol forcing from large eruptions are now linked not only to visible optical depth estimates on a global scale but also to details on the size, latitude and altitude distributions of the stratospheric aerosols. Based on our understanding of these uncertainties, we propose a new model-based approach to generating a volcanic forcing for general circulation model (GCM) and chemistry–climate model (CCM) simulations. This new volcanic forcing, covering the 1600–present period, uses an aerosol microphysical model to provide a realistic, physically consistent treatment of the stratospheric sulfate aerosols. Twenty-six eruptions were modeled individually using the latest available ice cores aerosol mass estimates and historical data on the latitude and date of eruptions. The evolution of aerosol spatial and size distribution after the sulfur dioxide discharge are hence characterized for each volcanic eruption. Large variations are seen in hemispheric partitioning and size distributions in relation to location/date of eruptions and injected SO2 masses. Results for recent eruptions show reasonable agreement with observations. By providing these new estimates of spatial distributions of shortwave and long-wave radiative perturbations, this volcanic forcing may help to better constrain the climate model responses to volcanic eruptions in the 1600–present period. The final data set consists of 3-D values (with constant longitude) of spectrally resolved extinction coefficients, single scattering albedos and asymmetry factors calculated for different wavelength bands upon request. Surface area densities for heterogeneous chemistry are also provided.

Galilee Blooming: First Palynological and Archaeological Data from an Early Byzantine Cistern at Horvat Kur

Preliminary archaeological and palynological results are presented from an early Byzantine cistern of the village Horvat Kur in eastern Lower Galilee/Israel. The rural site was settled from the Hellenistic until the Early Arab period, its synagogue was constructed shortly after 425 AD and renovated sometimes during the 2nd half of the 6th century AD. It was abandoned probably as a consequence of the earthquake of 749 AD. The intact and properly sealed cistern contained complete or fully restorable pottery. Two cooking pots from the early 5th century AD comprised sediments which was sampled for palynological purposes. Both samples, as well as a sample from the soil beneath one of the pots and a modern surface sample from the site, revealed well preserved palynomorphs in comparably high concentration showing a great potential of the cistern as a pollen archive. The pollen content points to an open, grassy semiarid landscape with an apparent scarcity of cultivars and trees in the vicinity of the site and an abundance of herbs, especially Asteraceae, which are still commonly found in modern regional vegetation.

Sudden Increase in Tidal Response Linked to Calving and Acceleration at a Large Greenland Outlet Glacier

Large calving events at Greenland's largest outlet glaciers are associated with glacial earthquakes and near instantaneous increases in glacier flow speed. At some glaciers and ice streams, flow is also modulated in a regular way by ocean tidal forcing at the terminus. At Helheim Glacier, analysis of geodetic data shows decimeter-level periodic position variations in response to tidal forcing. However, we also observe transient increases of more than 100% in the glacier's responsiveness to such tidal forcing following glacial-earthquake calving events. The timing and amplitude of the changes correlate strongly with the step-like increases in glacier speed and longitudinal strain rate associated with glacial earthquakes. The enhanced response to the ocean tides may be explained by a temporary disruption of the subglacial drainage system and a concomitant reduction of the friction at the ice-bedrock interface, and suggests a new means by which geodetic data may be used to infer glacier properties. Citation: de Juan, J., et al. (2010), Sudden increase in tidal response linked to calving and acceleration at a large Greenland outlet glacier, Geophys. Res. Lett., 37, L12501, doi: 10.1029/2010GL043289.