MALDI-TOF mass spectrometry for sub-typing of Streptococcus pneumoniae
Background Serotyping of Streptococcus pneumoniae is important for monitoring of vaccine impact. Unfortunately, conventional and molecular serotyping is expensive and technically demanding. This study aimed to determine the ability of matrix-assisted laser desorption-ionisation time-of-flight (MALDI...
| Главные авторы: | , , , , , , , , |
|---|---|
| Формат: | Journal article |
| Язык: | English |
| Опубликовано: |
BioMed Central
2020
|
| _version_ | 1826282382021361664 |
|---|---|
| author | Kann, S Sao, S Phoeung, C By, Y Bryant, J Komurian-Pradel, F Saphonn, V Chou, M Turner, P |
| author_facet | Kann, S Sao, S Phoeung, C By, Y Bryant, J Komurian-Pradel, F Saphonn, V Chou, M Turner, P |
| author_sort | Kann, S |
| collection | OXFORD |
| description | Background
Serotyping of Streptococcus pneumoniae is important for monitoring of vaccine impact. Unfortunately, conventional and molecular serotyping is expensive and technically demanding. This study aimed to determine the ability of matrix-assisted laser desorption-ionisation time-of-flight (MALDI-TOF) mass spectrometry to discriminate between pneumococcal serotypes and genotypes (defined by global pneumococcal sequence cluster, GPSC). In this study, MALDI-TOF mass spectra were generated for a diverse panel of whole genome sequenced pneumococcal isolates using the bioMerieux VITEK MS in clinical diagnostic (IVD) mode. Discriminatory mass peaks were identified and hierarchical clustering was performed to visually assess discriminatory ability. Random forest and classification and regression tree (CART) algorithms were used to formally determine how well serotypes and genotypes were identified by MALDI-TOF mass spectrum.
Results
One hundred and ninety-nine pneumococci, comprising 16 serotypes and non-typeable isolates from 46 GPSC, were analysed. In the primary experiment, hierarchical clustering revealed poor congruence between MALDI-TOF mass spectrum and serotype. The correct serotype was identified from MALDI-TOF mass spectrum in just 14.6% (random forest) or 35.4% (CART) of 130 isolates. Restricting the dataset to the nine dominant GPSC (61 isolates / 13 serotypes), discriminatory ability improved slightly: the correct serotype was identified in 21.3% (random forest) and 41.0% (CART). Finally, analysis of 69 isolates of three dominant serotype-genotype pairs (6B-GPSC1, 19F-GPSC23, 23F-GPSC624) resulted in the correct serotype identification in 81.1% (random forest) and 94.2% (CART) of isolates.
Conclusions
This work suggests that MALDI-TOF is not a useful technique for determination of pneumococcal serotype. MALDI-TOF mass spectra appear more associated with isolate genotype, which may still have utility for future pneumococcal surveillance activities.
|
| first_indexed | 2024-03-07T00:43:00Z |
| format | Journal article |
| id | oxford-uuid:83b0e79e-b429-4bad-90a9-0670d314cc86 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T00:43:00Z |
| publishDate | 2020 |
| publisher | BioMed Central |
| record_format | dspace |
| spelling | oxford-uuid:83b0e79e-b429-4bad-90a9-0670d314cc862022-03-26T21:45:56ZMALDI-TOF mass spectrometry for sub-typing of Streptococcus pneumoniaeJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:83b0e79e-b429-4bad-90a9-0670d314cc86EnglishSymplectic ElementsBioMed Central2020Kann, SSao, SPhoeung, CBy, YBryant, JKomurian-Pradel, FSaphonn, VChou, MTurner, PBackground Serotyping of Streptococcus pneumoniae is important for monitoring of vaccine impact. Unfortunately, conventional and molecular serotyping is expensive and technically demanding. This study aimed to determine the ability of matrix-assisted laser desorption-ionisation time-of-flight (MALDI-TOF) mass spectrometry to discriminate between pneumococcal serotypes and genotypes (defined by global pneumococcal sequence cluster, GPSC). In this study, MALDI-TOF mass spectra were generated for a diverse panel of whole genome sequenced pneumococcal isolates using the bioMerieux VITEK MS in clinical diagnostic (IVD) mode. Discriminatory mass peaks were identified and hierarchical clustering was performed to visually assess discriminatory ability. Random forest and classification and regression tree (CART) algorithms were used to formally determine how well serotypes and genotypes were identified by MALDI-TOF mass spectrum. Results One hundred and ninety-nine pneumococci, comprising 16 serotypes and non-typeable isolates from 46 GPSC, were analysed. In the primary experiment, hierarchical clustering revealed poor congruence between MALDI-TOF mass spectrum and serotype. The correct serotype was identified from MALDI-TOF mass spectrum in just 14.6% (random forest) or 35.4% (CART) of 130 isolates. Restricting the dataset to the nine dominant GPSC (61 isolates / 13 serotypes), discriminatory ability improved slightly: the correct serotype was identified in 21.3% (random forest) and 41.0% (CART). Finally, analysis of 69 isolates of three dominant serotype-genotype pairs (6B-GPSC1, 19F-GPSC23, 23F-GPSC624) resulted in the correct serotype identification in 81.1% (random forest) and 94.2% (CART) of isolates. Conclusions This work suggests that MALDI-TOF is not a useful technique for determination of pneumococcal serotype. MALDI-TOF mass spectra appear more associated with isolate genotype, which may still have utility for future pneumococcal surveillance activities. |
| spellingShingle | Kann, S Sao, S Phoeung, C By, Y Bryant, J Komurian-Pradel, F Saphonn, V Chou, M Turner, P MALDI-TOF mass spectrometry for sub-typing of Streptococcus pneumoniae |
| title | MALDI-TOF mass spectrometry for sub-typing of Streptococcus pneumoniae |
| title_full | MALDI-TOF mass spectrometry for sub-typing of Streptococcus pneumoniae |
| title_fullStr | MALDI-TOF mass spectrometry for sub-typing of Streptococcus pneumoniae |
| title_full_unstemmed | MALDI-TOF mass spectrometry for sub-typing of Streptococcus pneumoniae |
| title_short | MALDI-TOF mass spectrometry for sub-typing of Streptococcus pneumoniae |
| title_sort | maldi tof mass spectrometry for sub typing of streptococcus pneumoniae |
| work_keys_str_mv | AT kanns malditofmassspectrometryforsubtypingofstreptococcuspneumoniae AT saos malditofmassspectrometryforsubtypingofstreptococcuspneumoniae AT phoeungc malditofmassspectrometryforsubtypingofstreptococcuspneumoniae AT byy malditofmassspectrometryforsubtypingofstreptococcuspneumoniae AT bryantj malditofmassspectrometryforsubtypingofstreptococcuspneumoniae AT komurianpradelf malditofmassspectrometryforsubtypingofstreptococcuspneumoniae AT saphonnv malditofmassspectrometryforsubtypingofstreptococcuspneumoniae AT choum malditofmassspectrometryforsubtypingofstreptococcuspneumoniae AT turnerp malditofmassspectrometryforsubtypingofstreptococcuspneumoniae |