Group B streptococcus serotypes identification from MALDI-TOF mass spectra through data analysis and machine learning
| Autor: | Wen-Chi Li, 李文琪 |
|---|---|
| Rok vydání: | 2018 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 106 Group B streptococcus (GBS) is a commensal flora commonly found common in the natural environment and human gastrointestinal tract. GBS infection may cause sepsis or meningitis. The common routes of infection include direct transmission from infected mothers to infants, and transmission by contact. Besides, GBS may be also sexually transmitted. Serotyping of GBS is an essential tool in investigation of possible infection outbreak. Moreover, serotyping can facilitate identifying the source of infection, which is the key of efficient infection control. Although it is possible to serotype GBS through traditional biochemical experiments, it is a time-consuming and labor-intensive method. Matrix-assisted laser desorption ionization mass spectrometry is a technique that uses microbial cell protein composition to identify species. Its advantages are fast, low cost and high throughput. Therefore, this technique has been widely used for bacterial identification based on a specific protein profile. There are 325 mass spectra in this study by using this technique, including seven serotypes. In this work, a total of three features including peaks, peak pairs are investigated for correctly identifying the serotypes of GBS. And we used the machine learning and data mining techniques to construct five predictive models for different serotypes. It is hoped that medical personnel will be able to use the predictive model to more quickly identify the serotypes of GBS. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|