Publicação
An efficient and accurate framework for large-scale sequences of DNA barcodes
| Resumo: | DNA barcodes are short sequences of pre-defined gene regions that contain a sufficient amount of intra- and inter-species genetic information. High-throughput sequencing techniques are currently used to identify large sequences of DNA barcodes in a species genome, in a relatively short time. Domain experts require adequate self-contained tools to accurately and efficiently process DNA barcode data in a reasonable time, taking advantage of current parallel and heterogeneous computing systems. They also expect to use these tools on different computing platforms, from laptops to high-performance servers, without requiring a broad knowledge in software engineering to develop efficient computational applications. The main goal of this project was to develop a framework and associated user-friendly tools for domain experts to efficiently support DNA barcoding studies, providing an abstraction of the performance issues. 4SpecID is the key outcome of this work: an application software that integrates a semi-automated auditing and annotation tool for reference libraries, to ensure the quality standards of the compiled data, aiming to enable a grounded decision when identifying species from DNA barcodes. Its graphics interface aids the end user to specify the operations and it also simplifies data filtering and remote file handling. The C++ ported version (from MATLAB) was fully tested and is more robust than the original version. Architecture features common to laptop and compute servers were exploited, namely parallel programming techniques and memory models. The presented validation and performance results show significant improvements on execution times, not only on the sequential version, but also by using the available parallel capabilities of the underlying computing platforms. |
|---|---|
| Autores principais: | Neto, Luís Manuel Pacheco |
| Assunto: | High-throughput sequencing High performance computing DNA barcoding DNA metabarcodin Códigos de barras ADN Sequenciação de ADN Computação de alto desempenho |
| Ano: | 2021 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade do Minho |
| Idioma: | inglês |
| Origem: | RepositóriUM - Universidade do Minho |
| Resumo: | DNA barcodes are short sequences of pre-defined gene regions that contain a sufficient amount of intra- and inter-species genetic information. High-throughput sequencing techniques are currently used to identify large sequences of DNA barcodes in a species genome, in a relatively short time. Domain experts require adequate self-contained tools to accurately and efficiently process DNA barcode data in a reasonable time, taking advantage of current parallel and heterogeneous computing systems. They also expect to use these tools on different computing platforms, from laptops to high-performance servers, without requiring a broad knowledge in software engineering to develop efficient computational applications. The main goal of this project was to develop a framework and associated user-friendly tools for domain experts to efficiently support DNA barcoding studies, providing an abstraction of the performance issues. 4SpecID is the key outcome of this work: an application software that integrates a semi-automated auditing and annotation tool for reference libraries, to ensure the quality standards of the compiled data, aiming to enable a grounded decision when identifying species from DNA barcodes. Its graphics interface aids the end user to specify the operations and it also simplifies data filtering and remote file handling. The C++ ported version (from MATLAB) was fully tested and is more robust than the original version. Architecture features common to laptop and compute servers were exploited, namely parallel programming techniques and memory models. The presented validation and performance results show significant improvements on execution times, not only on the sequential version, but also by using the available parallel capabilities of the underlying computing platforms. |
|---|