Computational protocol: Genetic diversity and potential routes of transmission of Mycobacterium bovis in Mozambique

Similar protocols

Protocol publication

[…] In order to evaluate relatedness of the spoligotype and VNTR patterns of the different samples we generated a dendogram using the UPGMA algorithm using the tool provided on the site http://www.miru-vntrplus.org/. The spoligotype patterns of the respective isolates were entered into the Mbovis.org database. In this database each unique spoligotype pattern is named by 'SB' followed by a four integer number e.g. SB0120 []. To further evaluate the relatedness of the 178 spoligotypes and possible clusters within the data we generated a minimum spanning tree using the tool provided on the site http://www.MIRU-VNTRplus.org, and the individual allelic diversity was calculated for all 24 MIRU-VNTR loci using the same site. A maximum difference of 1 mutation within a Clonal Complex was considered for the definition of clusters.The index of discrimination [,] was calculated to determine the overall discriminatory power of the spoligotyping and MIRU-VNTR typing techniques using a tool provided on the site http://insilico.ehu.es/mini_tools/discriminatory_power. The number of isolates assigned to each type (spoligotyping and MIRU-VNTR) was introduced in the formula provided.The genotyping of the three extra VNTR markers for increasing the discrimination power of the analysis had the drawback that these three markers could not be used in the MIRU-VNTRplus website. A UPGMA tree was calculated in MEGA [] following a calculation of genetic distances between genotypes using the software Arlequin [].Phylogenetic reconstruction was performed on a larger dataset in order to contextualize the Mozambican diversity to the overall M. bovis diversity in Africa. In order to do so we collected data from 959 M. bovis strains (575 from Central and Western Africa, 104 from Eastern Africa, 6 from North Africa and 274 strains from Southern Africa) [,,,–]. The VNTR markers were very divergent between studies but 5 of these were present in most studies and for more than 90% of the samples (viz. MIRU2165, MIRU2461, MIRU577, MIRU580 and MIRU3192 or ETR-A, -B, -C, -D and -E). The set have been suggested to provide enough resolution in the African context []. To the data from these, we added the 43 spoligotyping markers. Sample selection was based on a strategy to maximize this resolution (43 spoligotyping markers and 5 VNTRs) and thus studies/samples lacking these markers were excluded. On that premise, all available African M. bovis data in the literature following these parameters was included, although some geographic areas are underrepresented (North Africa) compared to others (Central/Western Africa). Reconstruction was done using a matrix of these 48 markers and by applying the reduced median algorithm [] followed by the median joining algorithm [], both present at the network software (freely available at http://www.fluxus-engineering.com). The algorithm is able to reconstruct missing genotypes allowing some data to be included even if there are missing markers. However to keep this extrapolation to a minimum, only samples that missed a single VNTR marker were included. For comparison two M. tuberculosis samples and two M. caprae samples were extracted from the MIRU-VNTRplus database and included in the analysis. […]

Pipeline specifications

Software tools MIRU-VNTRplus, MEGA, Arlequin
Applications Phylogenetics, Population genetic analysis
Organisms Mycobacterium bovis, Mycoplasma bovis, Bos taurus, Homo sapiens