Machine learning predicts and provides insights into milk acidification rates of Lactococcus lactis

Screening approaches and potential of isolated lactic acid bacteria for improving fermentation of Saccharina latissima

BACKGROUND

With the growing interest in applying fermentation to seaweed biomasses, there is a need for fast and efficient selection of microbial strains that have the ability to 1) acidify quickly, 2) utilize seaweed constituents and c) exhibit some proteolytic activity. The present study aims to provide a fast methodology to screen large bacterial collections for potential applications in optimized seaweed fermentations, as well as investigate and assess the performance of a selected bacterial collection of the National Food Institute Culture Collection (NFICC) in seaweed fermentation. This approach is directed toward high-throughput (HT) methodologies, employing microwell assays for different phenotypical characteristics of lactic acid bacteria isolated from different sources. The overarching aim is the deeper understanding of the selection criteria when designing starter cultures for seaweed fermentation.

RESULTS

By employing high-throughput analytical workflows, the screening processing time is minimized, and among the different strains from a well-characterized strain collection, it was possible to distinguish between strong acidifiers and to replicate similar results when the volumes were scaled from 96-well plates to lab-scale fermentations (40 mL) of whole seaweed. Lactiplantibacillus plantarum, Lacticaseibacillus paracasei and, to a lesser extent, Lacticaseibacillus rhamnosus were among the fastest strains to reach the lowest endpoint pH values (< 4.5) in less than 48 h. Although the results regarding proteolytic capacity were not sufficient to prove that the candidates can also provide some flavor generation by the cleavage of proteins, NFICC1746 and NFICC2041 exhibited potential in releasing free alanine, glutamate and asparate as free amino acids.

CONCLUSIONS

With the described methodology, a large number of terrestrial lactic acid bacteria (LAB) isolates were screened for their performance and possible application for fermentation of brown sewaeeds. With a a fast conversion of sugars to organic acids, three potential new plant-isolated strains from NFICC, specifically Lactiplantibacillus plantarum ssp. argentoratensis (NFICC983), Lacticaseibacillus paracasei (NFICC1746) and Lacticaseibacillus rhamnosus (NFICC2041), were identified as promising candidates for future synthetic consortia aimed at application in bioprocessed seaweed. The combination of such strains will be the future focus to further optimize robust seaweed fermentations.

Modeling control and transduction of electrochemical gradients in acid-stressed bacteria

Transmembrane electrochemical gradients drive solute uptake and constitute a substantial fraction of the cellular energy pool in bacteria. These gradients act not only as "homeostatic contributors," but also play a dynamic and keystone role in several bacterial functions, including sensing, stress response, and metabolism. At the system level, multiple gradients interact with ion transporters and bacterial behavior in a complex, rapid, and emergent manner; consequently, experiments alone cannot untangle their interdependencies. Electrochemical gradient modeling provides a general framework to understand these interactions and their underlying mechanisms. We quantify the generation, maintenance, and interactions of electrical, proton, and potassium potential gradients under lactic acid-stress and lactic acid fermentation. Further, we elucidate a gradient-mediated mechanism for intracellular pH sensing and stress response. We demonstrate that this gradient model can yield insights on the energetic limitations of membrane transport, and can predict bacterial behavior across changing environments.

From genotype to phenotype: computational approaches for inferring microbial traits relevant to the food industry

When selecting microbial strains for the production of fermented foods, various microbial phenotypes need to be taken into account to achieve target product characteristics, such as biosafety, flavor, texture, and health-promoting effects. Through continuous advances in sequencing technologies, microbial whole-genome sequences of increasing quality can now be obtained both cheaper and faster, which increases the relevance of genome-based characterization of microbial phenotypes. Prediction of microbial phenotypes from genome sequences makes it possible to quickly screen large strain collections in silico to identify candidates with desirable traits. Several microbial phenotypes relevant to the production of fermented foods can be predicted using knowledge-based approaches, leveraging our existing understanding of the genetic and molecular mechanisms underlying those phenotypes. In the absence of this knowledge, data-driven approaches can be applied to estimate genotype-phenotype relationships based on large experimental datasets. Here, we review computational methods that implement knowledge- and data-driven approaches for phenotype prediction, as well as methods that combine elements from both approaches. Furthermore, we provide examples of how these methods have been applied in industrial biotechnology, with special focus on the fermented food industry.

A machine learning approach to explore individual risk factors for tuberculosis treatment non-adherence in Mukono district

Despite the availability and implementation of well-known efficacious interventions for tuberculosis treatment by the Ministry of Health, Uganda (MoH), treatment non-adherence persists. Moreover, identifying a specific tuberculosis patient at risk of treatment non-adherence is still a challenge. Thus, this retrospective study, based on a record review of 838 tuberculosis patients enrolled in six health facilities, presents, and discusses a machine learning approach to explore the individual risk factors predictive of tuberculosis treatment non-adherence in the Mukono district, Uganda. Five classification machine learning algorithms, logistic regression (LR), artificial neural networks (ANN), support vector machines (SVM), random forest (RF), and AdaBoost were trained, and evaluated by computing their accuracy, F1 score, precision, recall, and the area under the receiver operating curve (AUC) through the aid of a confusion matrix. Of the five developed and evaluated algorithms, SVM (91.28%) had the highest accuracy (AdaBoost, 91.05% performed better than SVM when AUC is considered as evaluation parameter). Looking at all five evaluation parameters globally, AdaBoost is quite on par with SVM. Individual risk factors predictive of non-adherence included tuberculosis type, GeneXpert results, sub-country, antiretroviral status, contacts below 5 years, health facility ownership, sputum test results at 2 months, treatment supporter, cotrimoxazole preventive therapy (CPT) dapsone status, risk group, patient age, gender, middle and upper arm circumference, referral, positive sputum test at 5 and 6 months. Therefore, machine learning techniques, specifically classification types, can identify patient factors predictive of treatment non-adherence and accurately differentiate between adherent and non-adherent patients. Thus, tuberculosis program management should consider adopting the classification machine learning techniques evaluated in this study as a screening tool for identifying and targeting suited interventions to these patients.