Benjamin M. Good

Tag clouds for summarizing web search results

In this paper, we describe an application, PubCloud that uses tagclouds for the summarization of results from queries over thePubMed database of biomedical literature. PubCloud responds toqueries of this database with tag clouds generated from wordsextracted from the abstracts returned by the query. The results ofa user study comparing the PubCloud tag-cloud summarization ofquery results with the standard result list provided by PubMedindicated that the tag cloud interface is advantageous in presenting descriptive information and in reducing user frustrationbut that it is less effective at the task of enabling users to discoverrelations between concepts.

Current V3 genotyping algorithms are inadequate for predicting X4 co-receptor usage in clinical isolates

Objective:Integrating CCR5 antagonists into clinical practice would benefit from accurate assays of co-receptor usage (CCR5 versus CXCR4) with fast turnaround and low cost. Design:Published HIV V3-loop based predictors of co-receptor usage were compared with actual phenotypic tropism results in a large cohort of antiretroviral naive individuals to determine accuracy on clinical samples and identify areas for improvement. Methods:Aligned HIV envelope V3 loop sequences (n = 977), derived by bulk sequencing were analyzed by six methods: the 11/25 rule; a neural network (NN), two support vector machines, and two subtype-B position specific scoring matrices (PSSM). Co-receptor phenotype results (Trofile Co-receptor Phenotype Assay; Monogram Biosciences) were stratified by CXCR4 relative light unit (RLU) readout and CD4 cell count. Results:Co-receptor phenotype was available for 920 clinical samples with V3 genotypes having fewer than seven amino acid mixtures (n = 769 R5; n = 151 X4-capable). Sensitivity and specificity for predicting X4 capacity were evaluated for the 11/25 rule (30% sensitivity/93% specificity), NN (44%/88%), PSSM(sinsi) (34%/96%), PSSM(x4r5) (24%/97%), SVMgenomiac (22%/90%) and SVMgeno2pheno (50%/89%). Quantitative increases in sensitivity could be obtained by optimizing the cut-off for methods with continuous output (PSSM methods), and/or integrating clinical data (CD4%). Sensitivity was directly proportional to strength of X4 signal in the phenotype assay (P < 0.05). Conclusions:Current default implementations of co-receptor prediction algorithms are inadequate for predicting HIV X4 co-receptor usage in clinical samples, particularly those X4 phenotypes with low CXCR4 RLU signals. Significant improvements can be made to genotypic predictors, including training on clinical samples, using additional data to improve predictions and optimizing cutoffs and increasing genotype sensitivity.

A new perspective on V3 phenotype prediction.

The particular coreceptor used by a strain of HIV-1 to enter a host cell is highly indicative of its pathology. HIV-1 coreceptor usage is primarily determined by the amino add sequences of the V3 loop region of the viral envelope glycoprotein. The canonical approach to sequence-based prediction of coreceptor usage was derived via statistical analysis of a less reliable and significantly smaller data set than is presently available. We aimed to produce a superior phenotypic classifier by applying modern machine learning (ML) techniques to the current database of V3 loop sequences with known phenotype. The trained classifiers along with the sequence data are available for public use at the supplementary website: http://genomiac2.ucsd.edu:8080/wetcat/v3.html and http://www.cs.waikato.ac.nz/ml/weka[corrected].

Genetic Composition of Human Immunodeficiency Virus Type 1 in Cerebrospinal Fluid and Blood without Treatment and during Failing Antiretroviral Therapy

ABSTRACT Human immunodeficiency virus (HIV) infection of the central nervous system (CNS) is a significant cause of morbidity. The requirements for HIV adaptation to the CNS for neuropathogenesis and the value of CSF virus as a surrogate for virus activity in brain parenchyma are not well established. We studied 18 HIV-infected subjects, most with advanced immunodeficiency and some neurocognitive impairment but none with evidence of opportunistic infection or malignancy of the CNS. Clonal sequences of C2-V3 env and population sequences of pol from HIV RNA in cerebrospinal fluid (CSF) and plasma were correlated with clinical and virologic variables. Most (14 of 18) subjects had partitioning of C2-V3 sequences according to compartment, and 9 of 13 subjects with drug resistance exhibited discordant resistance patterns between the two compartments. Regression analyses identified three to seven positions in C2-V3 that discriminated CSF from plasma HIV. The presence of compartmental differences at one or more of the identified positions in C2-V3 was highly associated with the presence of discordant resistance (P = 0.007), reflecting the autonomous replication of HIV and the independent evolution of drug resistance in the CNS. Discordance of resistance was associated with severity of neurocognitive deficits (P = 0.07), while low nadir CD4 counts were linked both to the severity of neurocognitive deficits and to discordant resistance patterns (P = 0.05 and 0.09, respectively). These observations support the study of CSF HIV as an accessible surrogate for HIV virions in the brain, confirm the high frequency of discordant resistance in subjects with advanced disease in the absence of opportunistic infection or malignancy of the CNS, and begin to identify genetic patterns in HIV env associated with adaptation to the CNS.

Semen-Specific Genetic Characteristics of Human Immunodeficiency Virus Type 1 env

ABSTRACT Human immunodeficiency virus type 1 (HIV-1) in the male genital tract may comprise virus produced locally in addition to virus transported from the circulation. Virus produced in the male genital tract may be genetically distinct, due to tissue-specific cellular characteristics and immunological pressures. HIV-1 env sequences derived from paired blood and semen samples from the Los Alamos HIV Sequence Database were analyzed to ascertain a male genital tract-specific viral signature. Machine learning algorithms could predict seminal tropism based on env sequences with accuracies exceeding 90%, suggesting that a strong genetic signature does exist for virus replicating in the male genital tract. Additionally, semen-derived viral populations exhibited constrained diversity (P < 0.05), decreased levels of positive selection (P < 0.025), decreased CXCR4 coreceptor utilization, and altered glycosylation patterns. Our analysis suggests that the male genital tract represents a distinct selective environment that contributes to the apparent genetic bottlenecks associated with the sexual transmission of HIV-1.

CIViC is a community knowledgebase for expert crowdsourcing the clinical interpretation of variants in cancer

CIViC is an expert-crowdsourced knowledgebase for Clinical Interpretation of Variants in Cancer describing the therapeutic, prognostic, diagnostic and predisposing relevance of inherited and somatic variants of all types. CIViC is committed to open-source code, open-access content, public application programming interfaces (APIs) and provenance of supporting evidence to allow for the transparent creation of current and accurate variant interpretations for use in cancer precision medicine.

Games with a scientific purpose.

The protein folding game Foldit shows that games are an effective way to recruit, engage and organize ordinary citizens to help solve difficult scientific problems.

Crowdsourcing in biomedicine: challenges and opportunities

The use of crowdsourcing to solve important but complex problems in biomedical and clinical sciences is growing and encompasses a wide variety of approaches. The crowd is diverse and includes online marketplace workers, health information seekers, science enthusiasts and domain experts. In this article, we review and highlight recent studies that use crowdsourcing to advance biomedicine. We classify these studies into two broad categories: (i) mining big data generated from a crowd (e.g. search logs) and (ii) active crowdsourcing via specific technical platforms, e.g. labor markets, wikis, scientific games and community challenges. Through describing each study in detail, we demonstrate the applicability of different methods in a variety of domains in biomedical research, including genomics, biocuration and clinical research. Furthermore, we discuss and highlight the strengths and limitations of different crowdsourcing platforms. Finally, we identify important emerging trends, opportunities and remaining challenges for future crowdsourcing research in biomedicine.

The Gene Wiki in 2011: community intelligence applied to human gene annotation

The Gene Wiki is an open-access and openly editable collection of Wikipedia articles about human genes. Initiated in 2008, it has grown to include articles about more than 10 000 genes that, collectively, contain more than 1.4 million words of gene-centric text with extensive citations back to the primary scientific literature. This growing body of useful, gene-centric content is the result of the work of thousands of individuals throughout the scientific community. Here, we describe recent improvements to the automated system that keeps the structured data presented on Gene Wiki articles in sync with the data from trusted primary databases. We also describe the expanding contents, editors and users of the Gene Wiki. Finally, we introduce a new automated system, called WikiTrust, which can effectively compute the quality of Wikipedia articles, including Gene Wiki articles, at the word level. All articles in the Gene Wiki can be freely accessed and edited at Wikipedia, and additional links and information can be found at the projects Wikipedia portal page: http://en.wikipedia.org/wiki/Portal:Gene_Wiki.

FAST, CHEAP AND OUT OF CONTROL: A ZERO CURATION MODEL FOR ONTOLOGY DEVELOPMENT

During two days at a conference focused on circulatory and respiratory health, 68 volunteers untrained in knowledge engineering participated in an experimental knowledge capture exercise. These volunteers created a shared vocabulary of 661 terms, linking these terms to each other and to a pre-existing upper ontology by adding 245 hyponym relationships and 340 synonym relationships. While ontology-building has proved to be an expensive and labor-intensive process using most existing methodologies, the rudimentary ontology constructed in this study was composed in only two days at a cost of only 3 t-shirts, 4 coffee mugs, and one chocolate moose. The protocol used to create and evaluate this ontology involved a targeted, web-based interface. The design and implementation of this protocol is discussed along with quantitative and qualitative assessments of the constructed ontology.