Development and application of bioinformatic tools for the representation and analysis of genetic diversity

Abstract

Genetic variation is the cornerstone of biological evolution. The description and explanation of the forces controlling genetic variation within and between populations is the main goal of population genetics. The deciphering of an explosive number of nucleotide sequences in different genes and species has changed radically the scope of population genetics, transforming it from an empirically insufficient science into a powerfully explanatory interdisciplinary endeavor, where high-throughput instruments generating new sequence data are integrated with bioinformatic tools for data mining and management, and advanced theoretical and statistical models for data interpretation. This thesis is an integrative and comprehensive bioinformatics and population genetics project whose central topic is the genetic diversity of populations. It is accomplished in three sequential steps: (i) the development of tools for data mining, processing, filtering and quality checking of raw data, (ii) the generation of databases of knowledge from refined data obtained in the first step, and (iii) the testing of hypotheses that require multi-species and/or multi-locus data. In the first part of the thesis, we have developed PDA Pipeline Diversity Analysis , an open-source, web-based tool that allows the exploration of polymorphism in large datasets of heterogeneous DNA sequences. This tool feeds from the millions of haplotypic sequences from individual studies that are stored in the main molecular biology databases, and generates high-quality, population genetics data that can be used to describe patterns of nucleotide variation in any species or gene. All the extracted and analyzed data resulting from the first part of this thesis is used in the second step to create a comprehensive on-line resource that provides searchable collections of polymorphic sequences with their associated diversity measures in the genus Drosophila (DPDB Drosophila Polymorphism Database ). This resource means an ambitious pledge to test the efficiency of the system created in the first part. Finally, two different studies that make use of the modules of data mining and analysis developed are shown. First, we study patterns of sequence variation to infer constraint and adaptation in Drosophila conserved noncoding sequences (CNSs). For this study we have used population genetics re-sequencing data from D. melanogaster together with comparative genomic data from other Drosophila species. We show that patterns of nucleotide sequence evolution in Drosophila CNSs are incompatible with the notion that mutational cold-spots explain these conserved blocks. Rather, the results support the hypothesis that CNSs are maintained by the action of purifying selection. The second study focuses on the coding evolution of Hox genes, a class of essential transcription factors expressed early in development that are involved in the specification of regional identities along the anteroposterior body axis. We have measured the rates of nucleotide divergence and fixation of insertions and deletions of three Hox genes, and compared them with those of three Hox-derived genes and a set non-Hox genes to test the hypothesis that Hox genes evolve slowly. Our results show that both the number of nonsynonymous substitutions and the degree of functional constraint are not significantly different between Hox and non-Hox genes, and that Hox and Hox-derived genes contain significantly more insertions and deletions than non-Hox genes in their coding sequences. Thus, Hox genes evolve faster than other essential genes expressed early in development, with complex expression patterns or with long introns rich in cis-regulatory elements. As a whole, the works presented in this thesis round a complete bioinformatics project off, including all the necessary steps from mining the data to generating new scientific knowledge. More interestingly, the outcome of each step is the seed of multiple possible studies in the next step, and thus this thesis has many applications for the scientific community.

Date of Award	20 Feb 2008
Original language	Undefined/Unknown
Supervisor	Antoni Barbadilla Prados (Director) & Alfredo Ruiz Panadero (Director)