| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 184.23 KB | Adobe PDF |
Advisor(s)
Abstract(s)
The increasing availability and sensitivity of analytical technologies has resulted in much higher complexity of molecular and cellular data, which biomarkers facilitate the bridge with a given biological state under scrutiny. Importantly, biomarkers are critical in spectroscopic technologies, as is the case of Fourier-transform infrared spectroscopy, for which different approaches to identify biomarkers have been established, for instance univariate statistical hypothesis tests. A workflow for the automatic application of said statistical tests is proposed with the objective of enabling a high throughput screening-approach that ensures statistical robustness, and works in a user-independent manner, thereby reducing analysis bias. The proposed workflow is the first step towards an automated method for biomarker screening, thus it is limited to distinguishing two populations. Firstly, the method used in previous work was improved to ensure the most powerful and accurate statistical tests are applied. Then, the statistical workflow was generalized to a more comprehensive range of data sets of two populations, for instance when sample size is not deemed high (less than thirty observations). Lastly, future work is outlined regarding the extension of the workflow to cases with more than two populations, but also of methods that can increase the pool of putative biomarkers that will then be screened for statistical significance.
Description
Este trabalho foi financiado pelo Concurso Anual para Projetos de Investigação, Desenvolvimento, Inovação e Criação Artística (IDI&CA) 2017 do Instituto Politécnico de Lisboa. Código de referência IPL/2017/DrugsPlatf/ISEL
Este trabalho foi financiado pelo Concurso Anual para Projetos de Investigação, Desenvolvimento, Inovação e Criação Artística (IDI&CA) 2018 do Instituto Politécnico de Lisboa. Código de referência IPL/2018/RENALPROG_ISEL
Este trabalho foi financiado pelo Concurso Anual para Projetos de Investigação, Desenvolvimento, Inovação e Criação Artística (IDI&CA) 2018 do Instituto Politécnico de Lisboa. Código de referência IPL/2018/RENALPROG_ISEL
Keywords
Workflow Biomarker screening Fourier-transform Infrared spectroscopy
Citation
CUNHA, Bernardo R.; [et al] – Towards an automated statistical workflow for biomarker screening in Fourier-transform infrared spectroscopy. In 2019 IEEE 6th Portuguese Meeting on Bioengineering (ENBENG). Lisbon, Portugal: IEEE, 2019. ISBN 978-1-5386-8506-8. Pp. 1-4
Publisher
Institute of Electrical and Electronics Engineers