Contents -- List of Figures -- List of Plates -- List of Tables -- 1 An introduction to text analysis in genomics -- 1.1 The genomics literature -- 1.2 Using text in genomics -- 1.2.1 Building databases of genetic knowledge -- 1.2.2 Analyzing experimental genomic data sets -- 1.2.3 Proposing new biological knowledge: identifying candidate genes -- 1.3 Publicly available text resources -- 1.3.1 Electronic text -- 1.3.2 Genome resources -- 1.3.3 Gene ontology -- 1.4 The advantage of text-based methods -- 1.5 Guide to this book -- 2 Functional genomics -- 2.1 Some molecular biology -- 2.1.1 Central dogma of molecular biology -- 2.1.2 Deoxyribonucleic acid -- 2.1.3 Ribonucleic acid -- 2.1.4 Genes -- 2.1.5 Proteins -- 2.1.6 Biological function -- 2.2 Probability theory and statistics -- 2.2.1 Probability -- 2.2.2 Conditional probability -- 2.2.3 Independence -- 2.2.4 Bayes' theorem -- 2.2.5 Probability distribution functions -- 2.2.6 Information theory -- 2.2.7 Population statistics -- 2.2.8 Measuring performance -- 2.3 Deriving and analyzing sequences -- 2.3.1 Sequencing -- 2.3.2 Homology -- 2.3.3 Sequence alignment -- 2.3.4 Pairwise sequence alignment and dynamic programming -- 2.3.5 Linear time pairwise alignment: BLAST -- 2.3.6 Multiple sequence alignment -- 2.3.7 Comparing sequences to profiles: weight matrices -- 2.3.8 Position specific iterative BLAST -- 2.3.9 Hidden Markov models -- 2.4 Gene expression profiling -- 2.4.1 Measuring gene expression with arrays -- 2.4.2 Measuring gene expression by sequencing and counting transcripts -- 2.4.3 Expression array analysis -- 2.4.4 Unsupervised grouping: clustering -- 2.4.5 k-means clustering -- 2.4.6 Self-organizing maps -- 2.4.7 Hierarchical clustering -- 2.4.8 Dimension reduction with principal components analysis.

2.4.9 Combining expression data with external information: supervised machine learning -- 2.4.10 Nearest neighbor classification -- 2.4.11 Linear discriminant analysis -- 3 Textual profiles of genes -- 3.1 Representing documents as word vectors -- 3.2 Metrics to compare documents -- 3.3 Some words are more important for document similarity -- 3.4 Building a vocabulary: feature selection -- 3.5 Weighting words -- 3.6 Latent semantic indexing -- 3.7 Defining textual profiles for genes -- 3.8 Using text like genomics data -- 3.9 A simple strategy to assigning keywords to groups of genes -- 3.10 Querying genes for biological function -- 4 Using text in sequence analysis -- 4.1 SWISS-PROT records as a textual resource -- 4.2 Using sequence similarity to extend literature references -- 4.3 Assigning keywords to summarize sequences hits -- 4.4 Using textual profiles to organize sequence hits -- 4.5 Using text to help identify remote homology -- 4.6 Modifying iterative sequence similarity searches to include text -- 4.7 Evaluating PSI-BLAST modified to include text -- 4.8 Combining sequence and text together -- 5 Text-based analysis of a single series of gene expression measurements -- 5.1 Pitfalls of gene expression analysis: noise -- 5.2 Phosphate metabolism: an example -- 5.3 The top fifteen genes -- 5.4 Distinguishing true positives from false positives with a literature-based approach -- 5.5 Neighbor expression information -- 5.6 Application to phosphate metabolism data set -- 5.7 Recognizing high induction false positives with literature-based scores -- 5.8 Recognizing low induction false positives -- 5.9 Assessing experiment quality with literature-based scoring -- 5.10 Improvements -- 5.11 Application to other assays -- 5.12 Assigning keywords that describe the broad biology of the experiment -- 6 Analyzing groups of genes.

6.1 Functional coherence of a group of genes -- 6.2 Overview of computational approach -- 6.3 Strategy to evaluate different algorithms -- 6.4 Word distribution divergence -- 6.5 Best article score -- 6.6 Neighbor divergence -- 6.6.1 Calculating a theoretical distribution of scores -- 6.6.2 Quantifying the difference between the empirical score distribution and the theoretical one -- 6.7 Neighbor divergence per gene -- 6.8 Corruption studies -- 6.9 Application of functional coherence scoring to screen gene expression clusters -- 6.10 Understanding the gene group's function -- 7 Analyzing large gene expression data sets -- 7.1 Groups of genes -- 7.2 Assigning keywords -- 7.3 Screening gene expression clusters -- 7.4 Optimizing cluster boundaries: hierarchical clustering -- 7.5 Application to other organisms besides yeast -- 7.6 Identifying and optimizing clusters in a Drosophila development data set -- 8 Using text classfication for gene function annotation -- 8.1 Functional vocabularies and gene annotation -- 8.1.1 Gene Ontology -- 8.1.2 Enzyme Commission -- 8.1.3 Kyoto Encyclopedia of Genes and Genomes -- 8.2 Text classification -- 8.3 Nearest neighbor classification -- 8.4 Naive Bayes classification -- 8.5 Maximum entropy classification -- 8.6 Feature selection: choosing the best words for classification -- 8.7 Classifying documents into functional categories -- 8.8 Comparing classifiers -- 8.9 Annotating genes -- 9 Finding gene names -- 9.1 Strategies to identify gene names -- 9.2 Recognizing gene names with a dictionary -- 9.3 Using word structure and appearance to identify gene names -- 9.4 Using syntax to eliminate gene name candidates -- 9.5 Using context as a clue about gene names -- 9.6 Morphology -- 9.7 Identifying gene names and their abbreviations -- 9.8. A single unified gene name finding algorithm -- 10 Protein interaction networks.

10.1 Genetic networks -- 10.2 Experimental assays to identify protein networks -- 10.2.1 Yeast two hybrid -- 10.2.2 Affinity precipitation -- 10.3 Predicting interactions versus verifying interactions with scientific text -- 10.4 Networks of co-occurring genes -- 10.5 Protein interactions and gene name co-occurrence in text -- 10.6 Number of textual co-occurrences predicts likelihood of an experimentally predicted interaction -- 10.7 Information extraction and genetic networks: increasing specificity and identifying interaction type -- 10.8 Statistical machine learning -- 11 Conclusion -- Index -- A -- B -- C -- D -- E -- F -- G -- H -- I -- J -- K -- L -- M -- N -- O -- P -- Q -- R -- S -- T -- U -- V -- W -- X -- Y -- Z.