Lamers SL, Salemi M, Beason S; International Conference on AIDS (15th : 2004 : Bangkok, Thailand).
Int Conf AIDS. 2004 Jul 11-16; 15: abstract no. TuPeA4378.
Gene Johnson Inc, St Augustine FL, United States
Background: HIVbase is software designed exclusively to help HIV scientists process genetic data. Accurate HIV alignments are necessary to answer questions about viral diversity and evolution, however, because of viral variation, high-quality HIV alignments are difficult to achieve. Scientists can spend large amounts of time hand-editing data in order to proceed with analysis. Two problems exist when aligning genetic sequences: speed and accuracy. The program Clustal is fast, but, for data that contains high genetic diversity, accuracy is often lost along the way. Our objective was to implement an object-oriented framework that provides a facility to develop specialized HIV alignment models. Clustal was written in a procedural fashion and does not lend itself easily to adaptation. Allowing various parts of the alignment process to be swapped out and replaced by components specific to a given virus genome would be more useful. Methods: Microsoft tools were used to write new algorithms, which were then built into HIVbase. V1-V2 sequences from various studies were isolated from a soon-to-be-released Web version of HIVbase. These domains are highly variable and contain significant length variation; therefore, these were optimal for testing the accuracy of the alignment. Here we compare HIV V1-V2 alignments created by both Clustal and HIVbase. Several methods are used to determine alignment accuracy (sight, phylogeny, intra-sample variation, conserved domains, secondary structure). Results: HIVbase provides an improved method for aligning highly variable HIV sequence genomes. The algorithm gives the user more flexibility throughout the different stages of the multiple alignment via a feature called round-tripping. The user can continue to realign the data set using separate processes for different regions of the alignment. For example, the scoring matrix, the substitution model, the pairwise alignment and clustering algorithms function independently. In other words, the alignment process in HIVbase is much more configurable than in current systems.
Publication Types:
Keywords:
- AIDS Vaccines
- Algorithms
- Base Sequence
- Cluster Analysis
- Genes, env
- Genome
- HIV
- HIV Envelope Protein gp120
- HIV-1
- Phylogeny
- Software
- genetics
- immunology
Other ID:
UI: 102282201
From Meeting Abstracts