The comparison between «low-molecular» and conventional approaches to species identification of bacteria by MALDI mass spectrometry
T13N4
B.L. Milman, V.V. Gostev, A.V. Dmitriev
The new approach to bacteria identification which uses the standard software for building mass spectral libraries of low-molecular compounds and for corresponding library searches, is compared to the conventional approach based on the commercial Biotyper software and database. The results are obtained for the random sample of 100 mass spectra of 25 strains of S. pyogenes, S. dysgalactiae subsp. equisimilis and S. anginosus. The spectra were sampled from the database of 728 mass spectra of 182 strains for some Streptococcus species. Both approaches were proved to result in the similar identification as 80–88 % of true outcomes. For similar reference mass spectra and the same identification criteria, the identification results were very close to each other as 24 agreements for 25 strains. This is the result from that different estimations of the mass spectral similarity included in programs under the comparison lead to the correlation of similarity indicators. A difference of identification results is mainly due to not the same reference databases and also different identification criteria.
The «low-molecular» approach to identification of microorganisms by MALDI mass spectrometry
T13N3
B.L. Milman, Yu.Yu. Ilyasov, N.V. Lugovkina, A.A. Golovina1, A.V. Dmitriev
The new approach to microorganism identification is presented which includes the transformation of their MALDI mass spectra directed to decreasing mass scale by the order of magnitude and the use of the standard software for building mass spectral libraries of low-molecular compounds and library searches. The library of 728 transformed mass spectra of 182 strains for some Streptococcus species was built. With this library, the rate of true identification was estimated by cross-comparison between library mass spectra. The true identification rate is 84 % for three Streptococcus species that corresponds to the average trueness of species identification by MALDI as found in the literature. The proposed approach to identification can be considered as the method of choice for solving problems of bacteria identification.
