Combining genetic algorithm with machine learning strategies for designing potent antimicrobial peptides.

BACKGROUND: Current methods in machine learning provide approaches for solving challenging, multiple constraint design problems. While deep learning and related neural networking methods have state-of-the-art performance, their vulnerability in decision making processes leading to irrational outcomes is a major concern for their implementation. With the rising antibiotic resistance, antimicrobial peptides (AMPs) have increasingly gained attention as novel therapeutic agents. This challenging design problem requires peptides which meet the multiple constraints of limiting drug-resistance in bacteria, preventing secondary infections from imbalanced microbial flora, and avoiding immune system suppression. AMPs offer a promising, bioinspired design space to targeting antimicrobial activity, but their versatility also requires the curated selection from a combinatorial sequence space. This space is too large for brute-force methods or currently known rational design approaches outside of machine learning. While there has been progress in using the design space to more effectively target AMP activity, a widely applicable approach has been elusive. The lack of transparency in machine learning has limited the advancement of scientific knowledge of how AMPs are related among each other, and the lack of general applicability for fully rational approaches has limited a broader understanding of the design space. METHODS: Here we combined an evolutionary method with rough set theory, a transparent machine learning approach, for designing antimicrobial peptides (AMPs). Our method achieves the customization of AMPs using supervised learning boundaries. Our system employs in vitro bacterial assays to measure fitness, codon-representation of peptides to gain flexibility of sequence selection in DNA-space with a genetic algorithm and machine learning to further accelerate the process. RESULTS: We use supervised machine learning and a genetic algorithm to find a peptide active against S. epidermidis, a common bacterial strain for implant infections, with an improved aggregation propensity average for an improved ease of synthesis. CONCLUSIONS: Our results demonstrate that AMP design can be customized to maintain activity and simplify production. To our knowledge, this is the first time when codon-based genetic algorithms combined with rough set theory methods is used for computational search on peptide sequences.

Antimicrobial peptide similarity and classification through rough set theory using physicochemical boundaries.

BACKGROUND: Antimicrobial peptides attract considerable interest as novel agents to combat infections. Their long-time potency across bacteria, viruses and fungi as part of diverse innate immune systems offers a solution to overcome the rising concerns from antibiotic resistance. With the rapid increase of antimicrobial peptides reported in the databases, peptide selection becomes a challenge. We propose similarity analyses to describe key properties that distinguish between active and non-active peptide sequences building upon the physicochemical properties of antimicrobial peptides. We used an iterative supervised machine learning approach to classify active peptides from inactive peptides with low false discovery rates in a relatively short computational search time. RESULTS: By generating explicit boundaries, our method defines new categories of active and inactive peptides based on their physicochemical properties. Consequently, it describes physicochemical characteristics of similarity among active peptides and the physicochemical boundaries between active and inactive peptides in a single process. To build the similarity boundaries, we used the rough set theory approach; to our knowledge, this is the first time that this approach has been used to classify peptides. The modified rough set theory method limits the number of values describing a boundary to a user-defined limit. Our method is optimized for specificity over selectivity. Noting that false positives increase activity assays while false negatives only increase computational search time, our method provided a low false discovery rate. Published datasets were used to compare our rough set theory method to other published classification methods and based on this comparison, we achieved high selectivity and comparable sensitivity to currently available methods. CONCLUSIONS: We developed rule sets that define physicochemical boundaries which allow us to directly classify the active sequences from inactive peptides. Existing classification methods are either sequence-order insensitive or length-dependent, whereas our method generates the rule sets that combine order-sensitive descriptors with length-independent descriptors. The method provides comparable or improved performance to currently available methods. Discovering the boundaries of physicochemical properties may lead to a new understanding of peptide similarity.

Computer-aided Molecular Design of Water Compatible Visible Light Photosensitizers for Dental Adhesive.

Dental adhesive resin undergoes phase separation during its infiltration through the wet demineralized dentin and it has been observed previously that the hydrophilic-rich phase is a vulnerable region for failure due to the lack of photo-polymerization and crosslinking density. The lack of photo-polymerization is mostly due to the partitioning of photo-initiators in low concentrations within this phase. Here, a computational approach has been employed to design candidate water compatible visible light photosensitizers which could improve the photo-polymerization of the hydrophilic-rich phase. This study is an extension of our previous work. QSPRs were developed for properties related to the photo-polymerization reaction of the adhesive monomers and hydrophilicity of the photosensitizer using connectivity indices as descriptors. QSPRs and structural constraints were formulated into an optimization problem which was solved stochastically via Tabu Search. Four candidate photosensitizer molecules have been proposed here which have the iminium ion as a common feature.

Effect of Partition of Photo-initiator Components and Addition of Iodonium Salt on the Photopolymerization of Phase-Separated Dental Adhesive.

The polymerization kinetics of physically separated hydrophobic- and hydrophilic-rich phases of a model dental adhesive have been investigated. The two phases were prepared from neat resin containing 2-hydroxyethyl methacrylate (HEMA) and bisphenol A glycerolate dimethacrylate (BisGMA) in the ratio of 45:55 (wt/wt). Neat resins containing various combinations of popular photo-initiating compounds, e.g., camphoquinone (CQ), ethyl 4-(dimethylamino)benzoate (EDMAB), 2-(dimethylamino)ethyl methacrylate (DMAEMA) and diphenyliodonium hexafluorophosphate (DPIHP) were prepared. To obtain the two phases 33 wt% of deuterium oxide (D2O) was added to the neat resins. This amount of D2O exceeded the miscibility limit for the resins. The concentration of each component of the photo-initiating system in the two phases was quantified by HPLC. When combined with CQ, DMAEMA is less efficient as a co-initiator compared to EDMAB. The addition of DPIHP as the third component into either CQ/EDMAB or CQ/DMAEMA photo-initiating systems leads to comparable performance in both the hydrophobic- and hydrophilic-rich phases. The addition of the iodonium salt significantly improved the photopolymerization of the hydrophilic-rich phase; the hydrophilic-rich phase exhibited extremely poor polymerization when the iodonium salt was not included in the formulation. The partition concentration of EDMAB in the hydrophilic-rich phase was significantly lower than that of DMAEMA or DPIHP. This study indicates the need for a combination of hydrophobic/hydrophilic photosensitizer and addition of iodonium salt to improve polymerization within the hydrophilic-rich phase of the dental adhesive.

Protein aggregation and lyophilization: Protein structural descriptors as predictors of aggregation propensity.

Lyophilization can induce aggregation in therapeutic proteins, but the relative importance of protein structure, formulation and processing conditions are poorly understood. To evaluate the contribution of protein structure to lyophilization-induced aggregation, fifteen proteins were co-lyophilized with each of five excipients. Extent of aggregation following lyophilization, measured using size-exclusion chromatography, was correlated with computational and biophysical protein structural descriptors via multiple linear regression. Descriptor selection was performed using exhaustive search and forward selection. The results demonstrate that, for a given excipient, extent of aggregation is highly correlated by eight to twelve structural descriptors. Leave-one-out cross validation showed that the correlations were able to successfully predict the aggregation for a protein "left out" of the data set. Selected descriptors varied with excipient, indicating both protein structure and excipient type contribute to lyophilization-induced aggregation. The results show some descriptors used to predict protein aggregation in solution are useful in predicting lyophilized protein aggregation.

Deconvoluting Kinetic Rate Constants of Catalytic Substrates from Scanning Electrochemical Approach Curves with Artificial Neural Networks.

Extracting information from experimental measurements in the chemical sciences typically requires curve fitting, deconvolution, and/or solving the governing partial differential equations via numerical (e.g., finite element analysis) or analytical methods. However, using numerical or analytical methods for high-throughput data analysis typically requires significant postprocessing efforts. Here, we show that deep learning artificial neural networks can be a very effective tool for extracting information from experimental data. As an example, reactivity and topography information from scanning electrochemical microscopy (SECM) approach curves are highly convoluted. This study utilized multilayer perceptrons and convolutional neural networks trained on simulated SECM data to extract kinetic rate constants of catalytic substrates. Our key findings were that multilayer perceptron models performed very well when the experimental data were close to the ideal conditions with which the model was trained. However, convolutional neural networks, which analyze images as opposed to direct data, were able to accurately predict the kinetic rate constant of Fe-doped nickel (oxy)hydroxide catalyst at different applied potentials even though the experimental approach curves were not ideal. Due to the speed at which machine learning models can analyze data, we believe this study shows that artificial neural networks could become powerful tools in high-throughput data analysis.

Use of glass transitions in carbohydrate excipient design for lyophilized protein formulations.

This work describes an effort to apply methods from process systems engineering to a pharmaceutical product design problem, with a novel application of statistical approaches to comparing solutions. A computational molecular design framework was employed to design carbohydrate molecules with high glass transition temperatures and low water content in the maximally freeze-concentrated matrix, with the objective of stabilizing lyophilized protein formulations. Quantitative structure-property relationships were developed for glass transition temperature of the anhydrous solute, glass transition temperature of the maximally concentrated solute, melting point of ice and Gordon-Taylor constant for carbohydrates. An optimization problem was formulated to design an excipient with optimal property values. Use of a stochastic optimization algorithm, Tabu search, provided several carbohydrate excipient candidates with statistically similar property values, as indicated by prediction intervals calculated for each property.

Impact of light intensity on the polymerization kinetics and network structure of model hydrophobic and hydrophilic methacrylate based dental adhesive resin.

The impact of light intensity on the degree of conversion (DC), rate of polymerization and network structure was investigated for hydrophobic and hydrophilic dental adhesive resins. Two and three component photoinitiating (PI) systems were used in this study. Low light intensities had a negative impact on the polymerization efficiency for the hydrophilic resin with 2 component PI system. Incorporation of iodonium salt in the hydrophilic resin significantly improved the polymerization efficiency of the HEMA/BisGMA system and led to a substantial DC, even at low light intensities. The results suggested that shorter polymer chains were formed in the presence of iodonium salt. It appears that there is little or no impact of light intensity on the polymer structure of the 2 component PI system. Light intensity has subtle impact on the polymer structure of the 3 component PI system. In the case of the hydrophobic resin, the polymer is so highly cross-linked that the presence of shorter chains for the 3 component PI system does not cause a decrease in the glass transition temperature (Tg ) when compared to the 2 component PI system. For the hydrophilic resin, the presence of shorter polymer chains in the 3 component PI system reduces the Tg when compared with the corresponding 2 component PI system. © 2015 Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater, 104B: 1666-1678, 2016.

Mathematical model of microbicidal flow dynamics and optimization of rheological properties for intra-vaginal drug delivery: Role of tissue mechanics and fluid rheology.

Topically applied microbicide gels can provide a self-administered and effective strategy to prevent sexually transmitted infections (STIs). We have investigated the interplay between vaginal tissue elasticity and the yield-stress of non-Newtonian fluids during microbicide deployment. We have developed a mathematical model of tissue deformation driven spreading of microbicidal gels based on thin film lubrication approximation and demonstrated the effect of tissue elasticity and fluid yield-stress on the spreading dynamics. Our results show that both elasticity of tissue and yield-stress rheology of gel are strong determinants of the coating behavior. An optimization framework has been demonstrated which leverages the flow dynamics of yield-stress fluid during deployment to maximize retention while reaching target coating length for a given tissue elasticity.

Adhesive/Dentin interface: the weak link in the composite restoration.

Results from clinical studies suggest that more than half of the 166 million dental restorations that were placed in the United States in 2005 were replacements for failed restorations. This emphasis on replacement therapy is expected to grow as dentists use composite as opposed to dental amalgam to restore moderate to large posterior lesions. Composite restorations have higher failure rates, more recurrent caries, and increased frequency of replacement as compared to amalgam. Penetration of bacterial enzymes, oral fluids, and bacteria into the crevices between the tooth and composite undermines the restoration and leads to recurrent decay and premature failure. Under in vivo conditions the bond formed at the adhesive/dentin interface can be the first defense against these noxious, damaging substances. The intent of this article is to review structural aspects of the clinical substrate that impact bond formation at the adhesive/dentin interface; to examine physico-chemical factors that affect the integrity and durability of the adhesive/dentin interfacial bond; and to explore how these factors act synergistically with mechanical forces to undermine the composite restoration. The article will examine the various avenues that have been pursued to address these problems and it will explore how alterations in material chemistry could address the detrimental impact of physico-chemical stresses on the bond formed at the adhesive/dentin interface.