Chun-Nan Hsu

FASTSNP: an always up-to-date and extendable service for SNP function analysis and prioritization

Single nucleotide polymorphism (SNP) prioritization based on the phenotypic risk is essential for association studies. Assessment of the risk requires access to a variety of heterogeneous biological databases and analytical tools. FASTSNP (function analysis and selection tool for single nucleotide polymorphisms) is a web server that allows users to efficiently identify and prioritize high-risk SNPs according to their phenotypic risks and putative functional effects. A unique feature of FASTSNP is that the functional effect information used for SNP prioritization is always up-to-date, because FASTSNP extracts the information from 11 external web servers at query time using a team of web wrapper agents. Moreover, FASTSNP is extendable by simply deploying more Web wrapper agents. To validate the results of our prioritization, we analyzed 1569 SNPs from the SNP500Cancer database. The results show that SNPs with a high predicted risk exhibit low allele frequencies for the minor alleles, consistent with a well-known finding that a strong selective pressure exists for functional polymorphisms. We have been using FASTSNP for 2 years and FASTSNP enables us to discover a novel promoter polymorphism. FASTSNP is available at .

Retrieving and Integrating Data from Multiple Information Sources

With the current explosion of data, retrieving and integrating information from various sources is a critical problem. Work in multidatabase systems has begun to address this problem, but it has primarily focused on methods for communicating between databases and requires significant effort for each new database added to the system. This paper describes a more general approach that exploits a semantic model of a problem domain to integrate the information from various information sources. The information sources handled include both databases and knowledge bases, and other information sources (e.g. programs) could potentially be incorporated into the system. This paper describes how both the domain and the information sources are modeled, shows how a query at the domain level is mapped into a set of queries to individual information sources, and presents algorithms for automatically improving the efficiency of queries using knowledge about both the domain and the information sources. This work is implemented in a system called SIMS and has been tested in a transportation planning domain using nine Oracle databases and a Loom knowledge base.

Generating finite-state transducers for semi-structured data extraction from the Web

Abstract Integrating a large number of Web information sources may significantly increase the utility of the World-Wide Web. A promising solution to the integration is through the use of a Web Information mediator that provides seamless, transparent access for the clients. Information mediators need wrappers to access a Web source as a structured database, but building wrappers by hand is impractical. Previous work on wrapper induction is too restrictive to handle a large number of Web pages that contain tuples with missing attributes, multiple values, variant attribute permutations, exceptions and typos. This paper presents SoftMealy, a novel wrapper representation formalism. This representation is based on a finite-state transducer (FST) and contextual rules. This approach can wrap a wide range of semistructured Web pages because FSTs can encode each different attribute permutation as a path. A SoftMealy wrapper can be induced from a handful of labeled examples using our generalization algorithm. We have implemented this approach into a prototype system and tested it on real Web pages. The performance statistics shows that the sizes of the induced wrappers as well as the required training effort are linear with regard to the structural variance of the test pages. Our experiment also shows that the induced wrappers can generalize over unseen pages.

Overview of BioCreative II gene mention recognition

Nineteen teams presented results for the Gene Mention Task at the BioCreative II Workshop. In this task participants designed systems to identify substrings in sentences corresponding to gene name mentions. A variety of different methods were used and the results varied with a highest achieved F1 score of 0.8721. Here we present brief descriptions of all the methods used and a statistical analysis of the results. We also demonstrate that, by combining the results from all submissions, an F score of 0.9066 is feasible, and furthermore that the best result makes use of the lowest scoring submissions.

The gene normalization task in BioCreative III

BackgroundWe report the Gene Normalization (GN) challenge in BioCreative III where participating teams were asked to return a ranked list of identifiers of the genes detected in full-text articles. For training, 32 fully and 500 partially annotated articles were prepared. A total of 507 articles were selected as the test set. Due to the high annotation cost, it was not feasible to obtain gold-standard human annotations for all test articles. Instead, we developed an Expectation Maximization (EM) algorithm approach for choosing a small number of test articles for manual annotation that were most capable of differentiating team performance. Moreover, the same algorithm was subsequently used for inferring ground truth based solely on team submissions. We report team performance on both gold standard and inferred ground truth using a newly proposed metric called Threshold Average Precision (TAP-k).ResultsWe received a total of 37 runs from 14 different teams for the task. When evaluated using the gold-standard annotations of the 50 articles, the highest TAP-k scores were 0.3297 (k=5), 0.3538 (k=10), and 0.3535 (k=20), respectively. Higher TAP-k scores of 0.4916 (k=5, 10, 20) were observed when evaluated using the inferred ground truth over the full test set. When combining team results using machine learning, the best composite system achieved TAP-k scores of 0.3707 (k=5), 0.4311 (k=10), and 0.4477 (k=20) on the gold standard, representing improvements of 12.4%, 21.8%, and 26.6% over the best team results, respectively.ConclusionsBy using full text and being species non-specific, the GN task in BioCreative III has moved closer to a real literature curation task than similar tasks in the past and presents additional challenges for the text mining community, as revealed in the overall team results. By evaluating teams using the gold standard, we show that the EM algorithm allows team submissions to be differentiated while keeping the manual annotation effort feasible. Using the inferred ground truth we show measures of comparative performance between teams. Finally, by comparing team rankings on gold standard vs. inferred ground truth, we further demonstrate that the inferred ground truth is as effective as the gold standard for detecting good team performance.

Weakly supervised learning of biomedical information extraction from curated data

BackgroundNumerous publicly available biomedical databases derive data by curating from literatures. The curated data can be useful as training examples for information extraction, but curated data usually lack the exact mentions and their locations in the text required for supervised machine learning. This paper describes a general approach to information extraction using curated data as training examples. The idea is to formulate the problem as cost-sensitive learning from noisy labels, where the cost is estimated by a committee of weak classifiers that consider both curated data and the text.ResultsWe test the idea on two information extraction tasks of Genome-Wide Association Studies (GWAS). The first task is to extract target phenotypes (diseases or traits) of a study and the second is to extract ethnicity backgrounds of study subjects for different stages (initial or replication). Experimental results show that our approach can achieve 87 % of Precision-at-2 (P@2) for disease/trait extraction, and 0.83 of F1-Score for stage-ethnicity extraction, both outperforming their cost-insensitive baseline counterparts.ConclusionsThe results show that curated biomedical databases can potentially be reused as training examples to train information extractors without expert annotation or refinement, opening an unprecedented opportunity of using “big data” in biomedical text mining.

Automatic information extraction from semi-structured Web pages by pattern discovery

The World Wide Web is now undeniably the richest and most dense source of information; yet, its structure makes it difficult to make use of that information in a systematic way. This paper proposes a pattern discovery approach to the rapid generation of information extractors that can extract structured data from semi-structured Web documents. Previous work in wrapper induction aims at learning extraction rules from user-labeled training examples, which, however, can be expensive in some practical applications. In this paper, we introduce IEPAD (an acronym for Information Extraction based on PAttern Discovery), a system that discovers extraction patterns from Web pages without user-labeled examples. IEPAD applies several pattern discovery techniques, including PAT-trees, multiple string alignments and pattern matching algorithms. Extractors generated by IEPAD can be generalized over unseen pages from the same Web data source. We empirically evaluate the performance of IEPAD on an information extraction task from 14 real Web data sources. Experimental results show that with the extraction rules discovered from a single page, IEPAD achieves 96% average retrieval rate, and with less than five example pages, IEPAD achieves 100% retrieval rate for 10 of the sample Web data sources.

The ANNIGMA-wrapper approach to fast feature selection for neural nets

This paper presents a novel feature selection approach for backpropagation neural networks (NNs). Previously, a feature selection technique known as the wrapper model was shown effective for decision trees induction. However, it is prohibitively expensive when applied to real-world neural net training characterized by large volumes of data and many feature choices. Our approach incorporates a weight analysis-based heuristic called artificial neural net input gain measurement approximation (ANNIGMA) to direct the search in the wrapper model and allows effective feature selection feasible for neural net applications. Experimental results on standard datasets show that this approach can efficiently reduce the number of features while maintaining or even improving the accuracy. We also report two successful applications of our approach in the helicopter maintenance applications.

The Use of Phenome-Wide Association Studies (PheWAS) for Exploration of Novel Genotype-Phenotype Relationships and Pleiotropy Discovery

The field of phenomics has been investigating network structure among large arrays of phenotypes, and genome‐wide association studies (GWAS) have been used to investigate the relationship between genetic variation and single diseases/outcomes. A novel approach has emerged combining both the exploration of phenotypic structure and genotypic variation, known as the phenome‐wide association study (PheWAS). The Population Architecture using Genomics and Epidemiology (PAGE) network is a National Human Genome Research Institute (NHGRI)‐supported collaboration of four groups accessing eight extensively characterized epidemiologic studies. The primary focus of PAGE is deep characterization of well‐replicated GWAS variants and their relationships to various phenotypes and traits in diverse epidemiologic studies that include European Americans, African Americans, Mexican Americans/Hispanics, Asians/Pacific Islanders, and Native Americans. The rich phenotypic resources of PAGE studies provide a unique opportunity for PheWAS as each genotyped variant can be tested for an association with the wide array of phenotypic measurements available within the studies of PAGE, including prevalent and incident status for multiple common clinical conditions and risk factors, as well as clinical parameters and intermediate biomarkers. The results of PheWAS can be used to discover novel relationships between SNPs, phenotypes, and networks of interrelated phenotypes; identify pleiotropy; provide novel mechanistic insights; and foster hypothesis generation. The PAGE network has developed infrastructure to support and perform PheWAS in a high‐throughput manner. As implementing the PheWAS approach has presented several challenges, the infrastructure and methodology, as well as insights gained in this project, are presented herein to benefit the larger scientific community. Genet. Epidemiol. 2011.

Integrating high dimensional bi-directional parsing models for gene mention tagging

Motivation: Tagging gene and gene product mentions in scientific text is an important initial step of literature mining. In this article, we describe in detail our gene mention tagger participated in BioCreative 2 challenge and analyze what contributes to its good performance. Our tagger is based on the conditional random fields model (CRF), the most prevailing method for the gene mention tagging task in BioCreative 2. Our tagger is interesting because it accomplished the highest F-scores among CRF-based methods and second over all. Moreover, we obtained our results by mostly applying open source packages, making it easy to duplicate our results. Results: We first describe in detail how we developed our CRF-based tagger. We designed a very high dimensional feature set that includes most of information that may be relevant. We trained bi-directional CRF models with the same set of features, one applies forward parsing and the other backward, and integrated two models based on the output scores and dictionary filtering. One of the most prominent factors that contributes to the good performance of our tagger is the integration of an additional backward parsing model. However, from the definition of CRF, it appears that a CRF model is symmetric and bi-directional parsing models will produce the same results. We show that due to different feature settings, a CRF model can be asymmetric and the feature setting for our tagger in BioCreative 2 not only produces different results but also gives backward parsing models slight but constant advantage over forward parsing model. To fully explore the potential of integrating bi-directional parsing models, we applied different asymmetric feature settings to generate many bi-directional parsing models and integrate them based on the output scores. Experimental results show that this integrated model can achieve even higher F-score solely based on the training corpus for gene mention tagging. Availability: Data sets, programs and an on-line service of our gene mention tagger can be accessed at http://aiia.iis.sinica.edu.tw/biocreative2.htm Contact: chunnan@iis.sinica.edu.tw