Search

This chapter introduces biological sequence analysis. Sequences arise in biological research because of the polymeric nature of the major biological macromolecules, nucleic acids, and proteins. The fundamental processes of molecular biology involves the capturing of the human genetic code in long molecules of DNA, which are packaged in the cell nucleus in the chromosomes. DNA is composed of a long chain of subunits—the nucleotides adenine, guanine, cytosine, and thymine (abbreviated A, G, C, and T respectively). DNA is usually found in the form of two paired chains, held together by hydrogen bonds between complementary pairs of nucleotides. A on one strand always pairs with T on the other, and similarly for G and C. Most of the important structural and functional components of a human cell are composed of proteins, which are long, folded chains composed of the 20 common amino acids, or peptides.

This study describes a new Hidden Markov Model (HMM) system for segmenting uncharacterized genomic DNA sequences into exons, introns, and intergenic regions. Separate HMM modules were designed and trained for specific regions of DNA: exons, introns, intergenic regions, and splice sites. The models were then tied together to form a biologically feasible topology. The integrated HMM was trained further on a set of eukaryotic DNA sequences and tested by using it to segment a separate set of sequences. The resulting HMM system which is called VEIL (Viterbi Exon-Intron Locator), obtains an overall accuracy on test data of 92% of total bases correctly labelled, with a correlation coefficient of 0.73. Using the more stringent test of exact exon prediction, VEIL correctly located both ends of 53% of the coding exons, and 49% of the exons it predicts are exactly correct. These results compare favorably to the best previous results for gene structure prediction and demonstrate the benefits of using HMMs for this problem.

The international Human Genome Project has entered a phase marked by phenomenal data growth. Rapid increases in the size and complexity of genomic maps, sequences, and macromolecular structures threaten to overwhelm the existing scientific information infrastructure. The traditional organization of public biological databases is inadequate to meet the problems arising from this increased flow. The Genome Database, an international repository of human genetic and physical mapping information, has been extensively revised to address these issues. Features of the new GDB architecture that were designed specifically to address expanding data volume include: direct public curation of the database, electronic data submission, efficient querying of specific genomic regions, and a modular database design. These features are described, with emphasis on their contribution to the query, display, and analysis of large data sets.

In 1991 the Genome Data Base at Johns Hopkins University School of Medicine was selected as the central repository for mapping data from the Human Genome Project, and was funded by NIH and DOE under a three year award. GDB has now finished 28 months of Federally funded operation. During this period a great deal of progress and many internal changes have taken place. In addition, many changes have also occurred in the external environment, and GDB has adapted its strategies to play an appropriate role in those changes as well. Recognizing the central role of mapping information in the genome project, it is important that GDB respond aggressively to the increasing demands of genomic researchers, as well as formulate a program of response to a number of long standing, but still unmet, needs of that community. It is even more important that GDB provide leadership in the genome informatics enterprise. Three themes described here are dominant in our future plans and represent the essence of the major changes made in the past year. They include: enhanced data acquisition, better map representation, and full integration into the collection of genomic databases.

Tactile perception of surface roughness was investigated in a combined psychophysical and neurophysiological study using 18 embossed, tetragonal dot patterns with varying dot spacings (1.3 to 6.2 mm) and sizes (0.4‐ to 1.2‐mm diameter). Psychophysical subjects explored these surfaces with the pad of the index finger and reported their estimates of subjective roughness. The results were inverted U‐shaped functions of spatial frequency. In neurophysiological experiments, exactly the same surfaces were scanned across the distal pads of macaque monkeys while recording from peripheral mechanoreceptive afferents (SA, QA, and PC afferents). Neural coding hypotheses were tested by comparing various features of the responses with the psychophysical results. Mean rate codes fared poorly as did codes based on “rate fluctuation” (summed absolute values of differences between rate maxima and minima). However, the second central moment of firing rate (variance) for the SA afferents matched the psychophysical data close...