Through advanced mechanistic modeling and the generation of large high-quality datasets, machine learning is becoming an integral part of understanding and engineering living systems. Here we show that mechanistic and machine learning models can be combined to enable accurate genotype-to-phenotype predictions. We use a genome-scale model to pinpoint engineering targets, efficient library construction of metabolic pathway designs, and high-throughput biosensor-enabled screening for training diverse machine learning algorithms. From a single data-generation cycle, this enables successful forward engineering of complex aromatic amino acid metabolism in yeast, with the best machine learning-guided design recommendations improving tryptophan titer and productivity by up to 74 and 43%, respectively, compared to the best designs used for algorithm training. Thus, this study highlights the power of combining mechanistic and machine learning models to effectively direct metabolic engineering efforts.

Practically all Bacillus thuringiensis strains contain a set of self-replicating, extrachromosomal DNA molecules or plasmids, which vary in number and size in the different strains. The plasmid patterns obtained from gel electrophoresis have previously been used as a tool to characterize strains, but comparison of the plasmid patterns has been limited in the number and diversity of strains analyzed. In this report, we were able to compare the plasmid patterns of 83 type strains (out of 84) and 47 additional strains from six serotypes. The information obtained from this comparison showed the importance of this tool as a strain characterization procedure and indicates the complexity and uniqueness of this feature. For example, with one exception, all type strains showed a unique plasmid pattern. All were unique in such a way that none showed even a single comigrating plasmid in the agarose gels, and therefore, cluster analysis was impossible, indicating that plasmid patterns are qualitative rather than quantitative features. Furthermore, comparison between strains belonging to the same serotype showed a great difference in variability. Some serotypes (e.g., israelensis) showed the same basic pattern among all its strains, while other serotypes (e.g., morrisoni) showed a great diversity of patterns. These results indicate that plasmid patterns are valuable tools to discriminate strains below the serotype level.Bacillus thuringiensis is a ubiquitous bacterium (25) isolated from a great diversity of habitats, which include soil, insects, stored crop products (31), phylloplane (32), and aquatic habitats (18). Its great biotechnological success resides in the production of highly specific insecticidal proteins (Cry proteins) simultaneously with the sporulation phase. These Cry proteins are coded by genes (cry genes) harbored in megaplasmids (10,14,21,34), although it has also been suggested that they are present in the chromosome (21). Plasmids have also been associated with the production of a different toxin called ␤-exotoxin (23). The relevance of plasmids in B. thuringiensis strains is assumed by the regular presence of a set of plasmids, which can vary in number from 1 to 17 and in size from 2 to 80 MDa (2, 9). The set of plasmids harbored in a strain are normally visualized by agarose gel electrophoresis, where they form an electrophoretic pattern of bands, according to their differential migration in the gel. Although B. thuringiensis plasmids have been studied either to locate cry genes (9, 21) or to transfer them between different strains and species (3,5,10,15,30), plasmid patterns have frequently been used to characterize strains (2, 29, 33), especially compared to those of standard strains (16,17,27). A plasmid pattern seems to be related to each strain, serotype, or any other subspecific group.In a plasmid pattern, two different groups of plasmids can be recognized: those that are Յ30 MDa and those that are Ն30 MDa, called megaplasmids. For practical purposes, each group is divided by the so-called chromos...

Bacillus is one of the main rhizobacteria to have been used as a study model for understanding many processes. However, their impact on photosynthetic metabolism has been poorly studied. The aim of this study was to evaluate the physiological parameters of pepper (Capsicum chinense Jacq.) plants inoculated with Bacillus spp. strains. Pepper seeds were inoculated with Bacillus cereus (K46 strain) and Bacillus spp. (M9 strain; a mixture of B. subtilis and B. amyloliquefaciens), chlorophyll fluorescence and gas exchange were evaluated. The ANOVA (P ≤ 0.05) showed that the maximum photochemical quantum yield of photosystem II (PSII) (F v /F m ) in plants inoculated with the M9 strain (0.784) increased with respect to other treatments (K46: 0.744 and Control: 0.739). Inoculated plants with M9 and K46 strains exhibited an increase of both photochemical quenching (qP) (by 27% and 24%, respectively) and CO 2 assimilation rate (photosynthesis) (by 20% and 16%, respectively), when compared with non-inoculated plants. Furthermore, plants inoculated with M9 and K46 showed decreased transpiration (61% and 57%, respectively) with respect to controls. Likewise, both electron transport rate of PSII (ETR) and PSII operating efficiency (Φ PSII ) increased in inoculated plants. However, only plants inoculated with the M9 strain showed enhancements on all growth characteristics. Our results therefore show that inoculating plants with M9 strain positively influenced the performance of the photosynthetic mechanism in pepper plants to increase chlorophyll fluorescence and gas exchange parameters. Promotion of photosynthetic capacity in pepper was due to increased ETR in the thylakoid membranes, which was promoted by the bacteria. M9 strain could even be used in sustainable agriculture programs.

Bacillus thuringiensis var israelensis was used to produce chitinase on shrimp wastes by fermentation at 30 °C and 250 rpm for 120 h. The enzyme was concentrated by ultrafiltration and was adjusted to pH 5.8. Antifungal chitinase activity on phytopathogenic fungi was investigated in growing cultures and on soybean seeds infested with Sclerotium rolfsii. Fungal inhibition was found to be 100% for S. rolfsii; 55% to 82% for A. terreus, A. flavus, Nigrospora sp, Rhizopus sp, A. niger, Fusarium sp, A. candidus,Absidia sp, and Helminthosporium sp; 45% for Curvularia sp; and 10% for A. fumigatus (P < 0.05). When soybean seeds were infected with S. rolfsii, germination was reduced from 93% to 25%; the addition of chitinase (0.8 U/mg protein) increased germination to 90%. B. thuringiensis chitinase may contribute to the biocontrol of S. rolfsii and other phytopathogenic fungi in soybean seeds in Integrated Pest Management programs.

Gas chromatography was utilized to determine triacylglycerol profiles in milk and non-milk fat. The values of triacylglycerol were subjected to linear discriminant analysis to detect and quantify non-milk fat in milk fat. Two groups of milk fat were analyzed: A) raw milk fat from the central region of Mexico (n = 216) and B) ultrapasteurized milk fat from 3 industries (n = 36), as well as pork lard (n = 2), bovine tallow (n = 2), fish oil (n = 2), peanut (n = 2), corn (n = 2), olive (n = 2), and soy (n = 2). The samples of raw milk fat were adulterated with non-milk fats in proportions of 0, 5, 10, 15, and 20% to form 5 groups. The first function obtained from the linear discriminant analysis allowed the correct classification of 94.4% of the samples with levels <10% of adulteration. The triacylglycerol values of the ultrapasteurized milk fats were evaluated with the discriminant function, demonstrating that one industry added non-milk fat to its product in 80% of the samples analyzed.