helveticus strains, and that our PCR-DGGE analysis provided a more accurate picture of the population composition of whey starters compared to culture-dependent techniques that often fail to isolate the most abundant strains. The production of aflatoxin (AF) B1 and B2 was determined during malting of wheat grains artificially contaminated with a toxigenic A. flavus strain (CCDCA 11553) isolated from craft beer raw material. Malting was performed in three steps (steeping, germination and kilning) following standard Central European Commission for Brewing Analysis procedures. AFB1 and AFB2 were quantified in eleven samples collected during the three malting steps and in malted wheat. Both, AFB1 and AFB2 were produced at the beginning of steeping and detected in all samples. The levels of AFB1 ranged from 229.35 to 455.66 μg/kg, and from 5.65 to 13.05 μg/kg for AFB2. The AFB2 increased during steeping, while no changes were observed in AFB1. Otherwise, AFB1 decreased during germination and AFB2 did not change. AFB1 and AFB2 increased after 16 h of kilning at 50 °C and decreased at the end of kilning, when the temperature reached 80 °C. The levels of AFB1 wheat malt were lower than those detected in wheat grains during steeping; however, levels of both AFB1 (240.46 μg/kg) and AFB2 (6.36 μg/kg) in Aspergillus flavus inoculated wheat malt exceeded the limits imposed by the regulatory agencies for cereals and derived products. In this work, the amplicon sequencing of the 16 S rRNA gene was employed to investigate the bacterial diversity in ingredients, processing environment, and ripened cheeses collected from three farms producing Serra da Canastra artisanal cheese. The data obtained indicated a remarkable variability in the bacteria consortia of the milk, whey, and environmental samples collected in farms 1, 2, and 3, despite their location in the same city. On the other hand, the starter culture and final product (ripened cheese) presented more constant and similar microbiota no matter the farm. The findings suggest that Streptococcus and Lactococcus have competitive advantages throughout Serra da Canastra cheese-making/ripening, which is crucial for their high relative abundance in the final products. An exploratory assessment based on sequencing data available in the literature showed that the Serra da Canastra cheeses sequences clustered with specific cheese varieties that are also made from raw milk but ripened for very different periods. The findings of this study highlight that despite the variability of milk and whey microbiota among the three farms, the starter culture ("pingo") has strong relevance in shaping the microbiota of the final product. The use of whole genome sequencing (WGS) data generated by short-read sequencing technologies such as the Illumina sequencing platforms has been shown to provide reliable results for Salmonella serotype prediction. Emerging long-read sequencing platforms developed by Oxford Nanopore Technologies (ONT) provide an alternative WGS method to meet the needs of industry for rapid and accurate Salmonella confirmation and serotype classification. Advantages of the ONT sequencing platforms include portability, real-time base-calling and long-read sequencing. To explore whether WGS data generated by an ONT sequencing platform could accurately predict Salmonella serotypes, 38 Salmonella strains representing 34 serotypes were sequenced using R9.4 flow cells on an ONT sequencer for up to 2 h. The downstream bioinformatics analysis was performed using pipelines with different assemblers including Canu, Wdbtg2 combined with Racon, or Miniasm combined with Racon. In silico serotype prediction programs were carried out using both SeqSero2 (raw reads and genome assemblies) and SISTR (genome assemblies). The WGS data of the same strains were also obtained from Illumina Hiseq (200 x depth of coverage per genome) as a benchmark of accurate serotype prediction. Predictions using WGS data generated after 30 min, 45 min, 1 h, and 2 h of ONT sequencing time all matched the prediction results from Illumina WGS data. This study demonstrated the comparable accuracy of WGS-based serotype prediction between ONT and Illumina sequencing platforms. This study also sets a start point for future validation of ONT WGS as a rapid Salmonella confirmation and serotype classification tool for the food industry. This study was conducted to address the dearth in works that simultaneously compare the growth and inactivation behaviors of selected pathogens in different milk products. In worst-case scenarios where hygienic practices are absent and heavy microbiological contaminations occur, Salmonella enterica, Escherichia coli O157H7, Listeria monocytogenes, Pseudomonas aeruginosa, and Staphylococcus aureus multiplied in all samples at room temperature (27 °C). Most organisms readily proliferated with growth lag (tlag) values ranging from 0.00 to 5.95 h.  https://www.selleckchem.com/products/sb290157-tfa.html  Growth rates (KG) ranged from 0.16 to 0.67 log CFU/h. Sanitary risk times (SRTs) for a 1-log population increase ranged from 1.85 to 6.27 h, while 3.69-12.55 h were the SRTs determined for 2-log population increase. Final populations (Popfin) ranged from 7.11 to 9.36 log CFU/mL. Inactivation in heavily contaminated milk during Holder pasteurization revealed biphasic inactivation behavior with total log reduction (TLR) after exposure to 62.5 °C for 30 min ranging from 1.91 (90.8%) to 6.00 (99.9999%). These results emphasize the importance food safety systems in the handling of milk and milk products during manufacture and preparation. Lactobacillus fermentum is a lactic acid bacterium frequently isolated from mammal tissues, milk, and plant material fermentations, such as sourdough. A comparative genomics analysis of 28 L. fermentum strains enabled the investigation of the core and accessory genes of this species. The core protein phylogenomic tree of the strains examined, consisting of five clades, did not exhibit clear clustering of strains based on isolation source, suggesting a free-living lifestyle. Based on the presence/absence of orthogroups, the largest clade, containing most of the human-related strains, was separated from the rest. The extended core genome included genes necessary for the heterolactic fermentation. Many traits were found to be strain-dependent, for instance utilisation of xylose and arabinose. Compared to other strains, the genome of L. fermentum IMDO 130101, a candidate starter culture strain capable of dominating sourdough fermentations, contained unique genes related to the metabolism of starch degradation products, which could be advantageous for growth in sourdough matrices.