Uncovering precision phenotype-biomarker associations in traumatic brain injury using topological data analysis.

BACKGROUND: Traumatic brain injury (TBI) is a complex disorder that is traditionally stratified based on clinical signs and symptoms. Recent imaging and molecular biomarker innovations provide unprecedented opportunities for improved TBI precision medicine, incorporating patho-anatomical and molecular mechanisms. Complete integration of these diverse data for TBI diagnosis and patient stratification remains an unmet challenge. METHODS AND FINDINGS: The Transforming Research and Clinical Knowledge in Traumatic Brain Injury (TRACK-TBI) Pilot multicenter study enrolled 586 acute TBI patients and collected diverse common data elements (TBI-CDEs) across the study population, including imaging, genetics, and clinical outcomes. We then applied topology-based data-driven discovery to identify natural subgroups of patients, based on the TBI-CDEs collected. Our hypothesis was two-fold: 1) A machine learning tool known as topological data analysis (TDA) would reveal data-driven patterns in patient outcomes to identify candidate biomarkers of recovery, and 2) TDA-identified biomarkers would significantly predict patient outcome recovery after TBI using more traditional methods of univariate statistical tests. TDA algorithms organized and mapped the data of TBI patients in multidimensional space, identifying a subset of mild TBI patients with a specific multivariate phenotype associated with unfavorable outcome at 3 and 6 months after injury. Further analyses revealed that this patient subset had high rates of post-traumatic stress disorder (PTSD), and enrichment in several distinct genetic polymorphisms associated with cellular responses to stress and DNA damage (PARP1), and in striatal dopamine processing (ANKK1, COMT, DRD2). CONCLUSIONS: TDA identified a unique diagnostic subgroup of patients with unfavorable outcome after mild TBI that were significantly predicted by the presence of specific genetic polymorphisms. Machine learning methods such as TDA may provide a robust method for patient stratification and treatment planning targeting identified biomarkers in future clinical trials in TBI patients. TRIAL REGISTRATION: ClinicalTrials.gov Identifier NCT01565551.

Topological data analysis for discovery in preclinical spinal cord injury and traumatic brain injury.

Data-driven discovery in complex neurological disorders has potential to extract meaningful syndromic knowledge from large, heterogeneous data sets to enhance potential for precision medicine. Here we describe the application of topological data analysis (TDA) for data-driven discovery in preclinical traumatic brain injury (TBI) and spinal cord injury (SCI) data sets mined from the Visualized Syndromic Information and Outcomes for Neurotrauma-SCI (VISION-SCI) repository. Through direct visualization of inter-related histopathological, functional and health outcomes, TDA detected novel patterns across the syndromic network, uncovering interactions between SCI and co-occurring TBI, as well as detrimental drug effects in unpublished multicentre preclinical drug trial data in SCI. TDA also revealed that perioperative hypertension predicted long-term recovery better than any tested drug after thoracic SCI in rats. TDA-based data-driven discovery has great potential application for decision-support for basic research and clinical problems such as outcome assessment, neurocritical care, treatment planning and rapid, precision-diagnosis.

Gene expression profile identifies tyrosine kinase c-Met as a targetable mediator of antiangiogenic therapy resistance.

PURPOSE: To identify mediators of glioblastoma antiangiogenic therapy resistance and target these mediators in xenografts. EXPERIMENTAL DESIGN: We conducted microarray analysis comparing bevacizumab-resistant glioblastomas (BRG) with pretreatment tumors from the same patients. We established novel xenograft models of antiangiogenic therapy resistance to target candidate resistance mediator(s). RESULTS: BRG microarray analysis revealed upregulation versus pretreatment of receptor tyrosine kinase c-Met, which underwent further investigation because of its prior biologic plausibility as a bevacizumab resistance mediator. BRGs exhibited increased hypoxia versus pretreatment in a manner correlating with their c-Met upregulation, increased c-Met phosphorylation, and increased phosphorylation of c-Met-activated focal adhesion kinase and STAT3. We developed 2 novel xenograft models of antiangiogenic therapy resistance. In the first model, serial bevacizumab treatment of an initially responsive xenograft generated a xenograft with acquired bevacizumab resistance, which exhibited upregulated c-Met expression versus pretreatment. In the second model, a BRG-derived xenograft maintained refractoriness to the MRI tumor vasculature alterations and survival-promoting effects of bevacizumab. Growth of this BRG-derived xenograft was inhibited by a c-Met inhibitor. Transducing these xenograft cells with c-Met short hairpin RNA inhibited their invasion and survival in hypoxia, disrupted their mesenchymal morphology, and converted them from bevacizumab-resistant to bevacizumab-responsive. Engineering bevacizumab-responsive cells to express constitutively active c-Met caused these cells to form bevacizumab-resistant xenografts. CONCLUSION: These findings support the role of c-Met in survival in hypoxia and invasion, features associated with antiangiogenic therapy resistance, and growth and therapeutic resistance of xenografts resistant to antiangiogenic therapy. Therapeutically targeting c-Met could prevent or overcome antiangiogenic therapy resistance.

Analysis of mRNA profiles after MEK1/2 inhibition in human pancreatic cancer cell lines reveals pathways involved in drug sensitivity.

Mutationally activated KRAS, detected in approximately 90% of pancreatic ductal adenocarcinomas (PDA), has proven an intractable pharmacologic target to date. Consequently, efforts to treat KRAS-mutated cancers are focused on targeting RAS-regulated signaling pathways. In mouse models, expression of BRAF(V600E) combined with dominant-negative TP53 elicits PDA, and pharmacologic blockade of mitogen-activated protein/extracellular signal-regulated kinase (MEK) inhibits proliferation of human PDA-derived cell lines. To better understand the role of RAF→MEK→ERK signaling on PDA cell proliferation, we assessed the consequences of MEK inhibition on global patterns of mRNA expression and tumor cell proliferation in a panel of human PDA-derived cell lines. This analysis revealed that RAF→MEK→ERK signaling regulates mRNAs involved in cell-cycle control as well as regulators of the immune system. Linear regression analysis of relative drug sensitivity and mRNA expression revealed mRNAs and pathways correlating with relative drug sensitivity of the cell lines. Mice carrying orthotopically implanted pancreas tumors that were treated with MEK inhibitor displayed reduced tumor growth, concomitant with a reduction of cells in S phase. Furthermore, analysis of tumor mRNA expression revealed PDA cell lines to display similar baseline and MEK inhibitor mRNA expression profiles in vitro and in vivo. Among the proteins subject to downregulation following MEK inhibition, we identified c-MYC as a key driver of cell proliferation downstream of RAF→MEK→ERK signaling. Indeed, in some PDA cell lines, RNA interference-mediated silencing of c-MYC expression had antiproliferative effects similar to that of MEK inhibition, thereby highlighting the importance of c-MYC in key aspects of pancreatic cancer cell maintenance.

An eQTL biological data visualization challenge and approaches from the visualization community.

In 2011, the IEEE VisWeek conferences inaugurated a symposium on Biological Data Visualization. Like other domain-oriented Vis symposia, this symposium's purpose was to explore the unique characteristics and requirements of visualization within the domain, and to enhance both the Visualization and Bio/Life-Sciences communities by pushing Biological data sets and domain understanding into the Visualization community, and well-informed Visualization solutions back to the Biological community. Amongst several other activities, the BioVis symposium created a data analysis and visualization contest. Unlike many contests in other venues, where the purpose is primarily to allow entrants to demonstrate tour-de-force programming skills on sample problems with known solutions, the BioVis contest was intended to whet the participants' appetites for a tremendously challenging biological domain, and simultaneously produce viable tools for a biological grand challenge domain with no extant solutions. For this purpose expression Quantitative Trait Locus (eQTL) data analysis was selected. In the BioVis 2011 contest, we provided contestants with a synthetic eQTL data set containing real biological variation, as well as a spiked-in gene expression interaction network influenced by single nucleotide polymorphism (SNP) DNA variation and a hypothetical disease model. Contestants were asked to elucidate the pattern of SNPs and interactions that predicted an individual's disease state. 9 teams competed in the contest using a mixture of methods, some analytical and others through visual exploratory methods. Independent panels of visualization and biological experts judged entries. Awards were given for each panel's favorite entry, and an overall best entry agreed upon by both panels. Three special mention awards were given for particularly innovative and useful aspects of those entries. And further recognition was given to entries that correctly answered a bonus question about how a proposed "gene therapy" change to a SNP might change an individual's disease status, which served as a calibration for each approaches' applicability to a typical domain question. In the future, BioVis will continue the data analysis and visualization contest, maintaining the philosophy of providing new challenging questions in open-ended and dramatically underserved Bio/Life Sciences domains.

Microarray analysis verifies two distinct phenotypes of glioblastomas resistant to antiangiogenic therapy.

PURPOSE: To identify mechanisms and mediators of resistance to antiangiogenic therapy in human glioblastoma. EXPERIMENTAL DESIGN: We carried out microarray gene expression analysis and immunohistochemistry comparing 21 recurrent glioblastomas progressing during antiangiogenic treatment with VEGF neutralizing antibody bevacizumab to paired pretreatment tumors from the same patients. RESULTS: Microarray analysis revealed that bevacizumab-resistant glioblastomas (BRG) had two clustering patterns defining subtypes that reflect radiographic growth patterns. Enhancing BRGs (EBRG) exhibited MRI enhancement, a long-established criterion for glioblastoma progression, and expressed mitogen-activated protein kinases, neural cell adhesion molecule-1 (NCAM-1), and aquaporin 4. Compared with their paired pretreatment tumors, EBRGs had unchanged vascularity and hypoxia, with increased proliferation. Nonenhancing BRGs (NBRG) exhibited minimal MRI enhancement but had FLAIR-bright expansion, a newer criterion for glioblastoma recurrence since the advent of antiangiogenic therapy, and expressed integrin α5, laminin, fibronectin1, and PDGFRß. NBRGs had less vascularity, more hypoxia, and unchanged proliferation than their paired pretreatment tumors. Primary NBRG cells exhibited more stellate morphology with a 3-fold increased shape factor and were nearly 4-fold more invasive in Matrigel chambers than primary cells from EBRGs or bevacizumab-naive glioblastomas (P < 0.05). CONCLUSION: Using microarray analysis, we found two resistance patterns during antiangiogenic therapy with distinct molecular profiles and radiographic growth patterns. These studies provide valuable biologic insight into the resistance that has limited antiangiogenic therapy to date.

Two distinct routes to oral cancer differing in genome instability and risk for cervical node metastasis.

PURPOSE: Problems in management of oral cancers or precancers include identification of patients at risk for metastasis, tumor recurrence, and second primary tumors or risk for progression of precancers (dysplasia) to cancer. Thus, the objective of this study was to clarify the role of genomic aberrations in oral cancer progression and metastasis. EXPERIMENTAL DESIGN: The spectrum of copy number alterations in oral dysplasia and squamous cell carcinomas (SCC) was determined by array comparative genomic hybridization. Associations with clinical characteristics were studied and results confirmed in an independent cohort. RESULTS: The presence of one or more of the chromosomal aberrations +3q24-qter, -8pter-p23.1, +8q12-q24.2, and +20 distinguishes a major subgroup (70%-80% of lesions, termed 3q8pq20 subtype) from the remainder (20%-30% of lesions, non-3q8pq20). The 3q8pq20 subtype is associated with chromosomal instability and differential methylation in the most chromosomally unstable tumors. The two subtypes differ significantly in clinical outcome with risk for cervical (neck) lymph node metastasis almost exclusively associated with the 3q8pq20 subtype in two independent oral SCC cohorts. CONCLUSIONS: Two subtypes of oral lesions indicative of at least two pathways for oral cancer development were distinguished that differ in chromosomal instability and risk for metastasis, suggesting that +3q,-8p, +8q, and +20 constitute a biomarker with clinical utility for identifying patients at risk for metastasis. Moreover, although increased numbers of genomic alterations can be harbingers of progression to cancer, dysplastic lesions lacking copy number changes cannot be considered benign as they are potential precursors to non-3q8pq20 locally invasive, yet not metastatic oral SCC.

EGAN: exploratory gene association networks.

UNLABELLED: Exploratory Gene Association Networks (EGAN) is a Java desktop application that provides a point-and-click environment for contextual graph visualization of high-throughput assay results. By loading the entire network of genes, pathways, interactions, annotation terms and literature references directly into memory, EGAN allows a biologist to repeatedly query and interpret multiple experimental results without incurring additional delays for data download/integration. Other compelling features of EGAN include: support for diverse -omics technologies, a simple and interactive graph display, sortable/searchable data tables, links to external web resources including > or = 240 000 articles at PubMed, hypergeometric and GSEA-like enrichment statistics, pipeline-compatible automation via scripting and the ability to completely customize and/or supplement the network with new/proprietary data. AVAILABILITY: Runs on most operating systems via Java; downloadable from http://akt.ucsf.edu/EGAN/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Concept recognition for extracting protein interaction relations from biomedical text.

BACKGROUND: Reliable information extraction applications have been a long sought goal of the biomedical text mining community, a goal that if reached would provide valuable tools to benchside biologists in their increasingly difficult task of assimilating the knowledge contained in the biomedical literature. We present an integrated approach to concept recognition in biomedical text. Concept recognition provides key information that has been largely missing from previous biomedical information extraction efforts, namely direct links to well defined knowledge resources that explicitly cement the concept's semantics. The BioCreative II tasks discussed in this special issue have provided a unique opportunity to demonstrate the effectiveness of concept recognition in the field of biomedical language processing. RESULTS: Through the modular construction of a protein interaction relation extraction system, we present several use cases of concept recognition in biomedical text, and relate these use cases to potential uses by the benchside biologist. CONCLUSION: Current information extraction technologies are approaching performance standards at which concept recognition can begin to deliver high quality data to the benchside biologist. Our system is available as part of the BioCreative Meta-Server project and on the internet http://bionlp.sourceforge.net.

Identification of antibiotic stress-inducible promoters: a systematic approach to novel pathway-specific reporter assays for antibacterial drug discovery.

As present antibiotics therapy becomes increasingly ineffectual, new technologies are required to identify and develop novel classes of antibacterial agents. An attractive alternative to the classical target-based approach is the use of promoter-inducible reporter assays for high-throughput screening. The wide usage of these assays is, however, limited by the small number of specifically responding promoters that are known at present. This work describes a novel approach for identifying genetic regulators that are suitable for the design of pathway-specific assays. The basis for the proposed strategy is a large set of antibiotics-triggered expression profiles ("Reference Compendium"). Pattern recognition algorithms applied to the expression data pinpoint the relevant transcription-factor-binding sites in whole-genome sequences. Using this technique, we constructed a fatty-acid-pathway-specific reporter assay that is based on a novel stress-inducible promoter. In a proof-of-principle experiment, this assay was shown to enable screening for new small-molecule inhibitors of bacterial growth.