In order to ensure the quality of raw milk and enrich quality control strains for the acceptance evaluation of culture media for the detection of coliform bacteria in raw milk, three brands of violet red bile agar (VRBA), violet red bile agar with 4-methylumbelliferyl-β-D-glucuronidehydrate (VRBA-MUG), lauryl tryptose broth (LST) and brilliant green lactose bile broth (BGLB) were selected and evaluated for their physical indicators such as appearance, pH and moisture content as well as microbial indicators such as growth rate and selectivity using the methods specified in China’s national standard for quality requirements for culture media and reagents in food microbiological examination (GB 4789.28-2013). Besides, using Citrobacter freundii ATCC 43864 as a reference strain, seven strains of C. freundii were selected for acceptance evaluation of the four media. The results showed that the color of all three brands of dry powdered media was normal. Their moisture content was 5%–6%, and their pH was in line with the national standard. The growth rate of the seven tested strains of C. freundii was greater than 0.7 on the three brands of VRBA and VRBA-MUG, while only CMCC 48098 and CICC 22910 achieved turbidity of 2 on the three brands of LST and BGLB, and produced gas in the catheter; the same results were obtained with ATCC 43864. In summary, the three brands of VRBA, VRBA-MUG, LST and BGLB all comply with GB 4789.28-2013, and CMCC 48098 and CICC 22910 can be recommended as supplements to quality control strains for the acceptance evaluation of the four media.

An uncertainty evaluation method for the microbiological determination of biotin in infant formula was developed in order to control the quality of test results and improve the detection accuracy, a method for uncertainty evaluation of milk powder by method was established. According to the requirements of the Evaluation and Expression of Uncertainty in Measurement (JJF 1059.1—2012) and the Guide to the Evaluation and Expression of Uncertainty in Measurement (GB/T 27418—2017), the uncertainty components were estimated for the determination of biotin in infant formula by the method specified in the National Food Safety Standard Determination of Biotin in Foods (GB 5009.259—2016). It was found that the greatest uncertainty resulted from standard curve fitting and repeatability test. The uncertainty for the biotin content of three infant formula samples was as follows: sample A (19.87 ± 2.84) μg/100 g, B (30.35 ± 4.41) μg/100 g, and C (53.15 ± 7.15) μg/100 g (in the 95% confidence interval, coverage factor (k) = 2). As the content of biotin in samples increased, the expanded uncertainty also increased.

A droplet digital polymerase chain reaction (ddPCR) method for the quantitative detection Acetobacter aceti in fermented milk was established. Specific primers and probes were designed according to the internally transcribed spacer (ITS) gene sequence of Acetobacter aceti, and annealing temperature was optimized. The specificity of the method was verified by applying it on various strains, the limit of detection (LOD) was determined for artificially inoculated Acetobacter aceti, and the absolute quantification was systematically investigated by comparing the results of ddPCR and the counting results. The experimental results showed that the optimal annealing temperature was 54.6 ℃, the method had strong specificity and high sensitivity, and the LOD was 7.2 × 101 CFU/mL. The quantitative deviation rate was 23.73%. This method can meet the demand for quantitative detection of Acetobacter aceti in fermented milk.

A rapid detection method based on helicase-dependent isothermal DNA amplification (HDA) was established for Lactobacillus plantarum in fermented milk. Specific primers were designed according to the scrB gene sequence of Lactobacillus plantarum (Genbank Accession NO. AJ579541.1), and the optimal concentrations of UvrD helicase and T4 gp32 in the reaction system were determined through experiments. The limit of detection (LOD), specificity, consistency and stability of the proposed method were evaluated by use of L. plantarum-spiked samples, amplification of various strains, and electrophoresis of amplified products and sequence alignment analysis, respectively. The results showed that the optimized of UvrD helicase and T4 gp32 in the reaction system were found to be 0.15 and 5.0 μg, respectively. The HDA method had high specificity with no amplification of other strains tested. The detection limit was 2.8 × 101 CFU/g. The amplified product was consistent with the designed sequence length (273 bp) and the sequence homology was 100%. In conclusion: this method is rapid, simple, sensitive and suitable for the detection of Lactobacillus plantarum in fermented milk.

A quality control sample for the qualitative analysis of Listeria monocytogenes was developed by vacuum freezing drying technology, and the homogeneity and stability of the quality control sample were systematically analyzed. By optimizing the type of lyophilized matrix, the optimal conditions for the development of quality control samples were obtained. The homogeneity and stability of the developed quality control sample were verified by counting the number of cells as well as using matrix-assisted laser desorption ionization time-of-flight mass spectrometry. The quality control sample was white in color, spherical in shape and uniform in size. The results of cell counting for uniformity validation showed F = 0.567, which was less than the critical value, indicating good uniformity. In the transportation stability test, the quality control sample remained stable at 37 and 25 ℃. In the storage stability test, the resurrection rate of the quality control sample was 101.5% after 28 days of storage at ?20 ℃, and 99.6% after 28 days of storage at 4 ℃, indicating that the sample has good uniformity and stability, and can be used as a positive quality control sample for the detection and quality control of Listeria monocytogenes.

In order to evaluate the application value of ready-to-use Salmonella quality control samples in matrix assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) identification of Salmonella in milk-containing formula foods for special medical purposes, this experiment evaluated the uniformity and stability of the quality control samples. After being stored at 20 ℃ for 0, 14 and 28 days, these samples were identified by mass spectrometry and the VITEK automatic microbial analysis system. The results showed that the quality control samples had good stability and uniformity, and their mass spectra were similar to each other. The VITEK system confirmed their identity as Salmonella. In conclusion, the ready-to-use Salmonella quality control strains can be used for the detection of Salmonella in milk-containing formula foods for special medical purposes.