Machine learning and genetic algorithm-guided directed evolution for the development of antimicrobial peptides.

INTRODUCTION: Antimicrobial peptides (AMPs) are valuable alternatives to traditional antibiotics, possess a variety of potent biological activities and exhibit immunomodulatory effects that alleviate difficult-to-treat infections. Clarifying the structure-activity relationships of AMPs can direct the synthesis of desirable peptide therapeutics. OBJECTIVES: In this study, the lipopolysaccharide-binding domain (LBD) was identified through machine learning-guided directed evolution, which acts as a functional domain of the anti-lipopolysaccharide factor family of AMPs identified from Marsupenaeus japonicus. METHODS: LBDA-D was identified as an output of this algorithm, in which the original LBDMj sequence was the input, and the three-dimensional solution structure of LBDB was determined using nuclear magnetic resonance. Furthermore, our study involved a comprehensive series of experiments, including morphological studies and in vitro and in vivo antibacterial tests. RESULTS: The NMR solution structure showed that LBDB possesses a circular extended structure with a disulfide crosslink at the terminus and two 310-helices and exhibits a broad antimicrobial spectrum. In addition, scanning electron microscopy (SEM) and transmission electron microscopy (TEM) showed that LBDB induced the formation of a cluster of bacteria wrapped in a flexible coating that ruptured and consequently killed the bacteria. Finally, coinjection of LBDB, Vibrio alginolyticus and Staphylococcus aureus in vivo improved the survival of M. japonicus, demonstrating the promising therapeutic role of LBDB for treating infectious disease. CONCLUSIONS: The findings of this study pave the way for the rational drug design of activity-enhanced peptide antibiotics.

The effect of microplastics on the depuration of hydrophobic organic contaminants in Daphnia magna: A quantitative model analysis.

Microplastics are emerging pollutants that can absorb large amounts of hydrophobic organic contaminants (HOCs). However, no biodynamic model has yet been proposed to estimate their effects on HOC depuration in aquatic organisms, where the HOC concentrations are time-varying. In this work, a microplastic-inclusive biodynamic model was developed to estimate the depuration of HOCs via ingestion of microplastics. Several key parameters of the model were redefined to determine the dynamic HOC concentrations. Through the parameterized model, the relative contributions of dermal and intestinal pathways can be distinguished. Moreover, the model was verified and the vector effect of microplastics was confirmed by studying the depuration of polychlorinated biphenyl (PCB) in Daphnia magna (D. magna) with different sizes of polystyrene (PS) microplastics. The results showed that microplastics contributed to the elimination kinetics of PCBs because of the fugacity gradient between the ingested microplastics and the biota lipids, especially for the less hydrophobic PCBs. The intestinal elimination pathway via microplastics would promote overall PCB elimination, contributing 37-41 % and 29-35 % to the total flux in the 100 nm and 2 µm polystyrene (PS) microplastic suspensions, respectively. Furthermore, the contribution of microplastic uptake to total HOC elimination increased with decreasing microplastic size in water, suggesting that microplastics may protect organisms from HOC risks. In conclusion, this work demonstrated that the proposed biodynamic model is capable of estimating the dynamic depuration of HOCs for aquatic organisms. The results can shed light on a better understanding of the vector effects of microplastics.

Dynamic logistic state space prediction model for clinical decision making.

Prediction modeling for clinical decision making is of great importance and needed to be updated frequently with the changes of patient population and clinical practice. Existing methods are either done in an ad hoc fashion, such as model recalibration or focus on studying the relationship between predictors and outcome and less so for the purpose of prediction. In this article, we propose a dynamic logistic state space model to continuously update the parameters whenever new information becomes available. The proposed model allows for both time-varying and time-invariant coefficients. The varying coefficients are modeled using smoothing splines to account for their smooth trends over time. The smoothing parameters are objectively chosen by maximum likelihood. The model is updated using batch data accumulated at prespecified time intervals, which allows for better approximation of the underlying binomial density function. In the simulation, we show that the new model has significantly higher prediction accuracy compared to existing methods. We apply the method to predict 1 year survival after lung transplantation using the United Network for Organ Sharing data.

Modified Brier score for evaluating prediction accuracy for binary outcomes.

The Brier score has been a popular measure of prediction accuracy for binary outcomes. However, it is not straightforward to interpret the Brier score for a prediction model since its value depends on the outcome prevalence. We decompose the Brier score into two components, the mean squares between the estimated and true underlying binary probabilities, and the variance of the binary outcome that is not reflective of the model performance. We then propose to modify the Brier score by removing the variance of the binary outcome, estimated via a general sliding window approach. We show that the new proposed measure is more sensitive for comparing different models through simulation. A standardized performance improvement measure is also proposed based on the new criterion to quantify the improvement of prediction performance. We apply the new measures to the data from the Breast Cancer Surveillance Consortium and compare the performance of predicting breast cancer risk using the models with and without its most important predictor.