Discovering knowledge from medical databases using evolutionary algorithms

Multiobjective grammar-based genetic programming applied to the study of asthma and allergy epidemiology

Background

Asthma and allergies prevalence increased in recent decades, being a serious global health problem. They are complex diseases with strong contextual influence, so that the use of advanced machine learning tools such as genetic programming could be important for the understanding the causal mechanisms explaining those conditions. Here, we applied a multiobjective grammar-based genetic programming (MGGP) to a dataset composed by 1047 subjects. The dataset contains information on the environmental, psychosocial, socioeconomics, nutritional and infectious factors collected from participating children. The objective of this work is to generate models that explain the occurrence of asthma, and two markers of allergy: presence of IgE antibody against common allergens, and skin prick test positivity for common allergens (SPT).

Results

The average of the accuracies of the models for asthma higher in MGGP than C4.5. IgE were higher in MGGP than in both, logistic regression and C4.5. MGGP had levels of accuracy similar to RF, but unlike RF, MGGP was able to generate models that were easy to interpret.

Conclusions

MGGP has shown that infections, psychosocial, nutritional, hygiene, and socioeconomic factors may be related in such an intricate way, that could be hardly detected using traditional regression based epidemiological techniques. The algorithm MGGP was implemented in c ++ and is available on repository: http://bitbucket.org/ciml-ufjf/ciml-lib.

Electronic supplementary material

The online version of this article (10.1186/s12859-018-2233-z) contains supplementary material, which is available to authorized users.

Competitive binding assay for glucose based on glycodendrimer-fluorophore conjugates

A new fluorescent glucose assay has been created using Alexa Fluor 647-labeled concanavalin A (Con A) and a fourth-generation PAMAM Alexa Fluor 594-labeled glycodendrimer. This assay has been shown to have a large response to glucose within the biological range and to be capable of functioning within a polymer hydrogel. In this paper, the glucose response is shown to be a single fluorophore-based quenching reaction. Data showing that the sensor is fully reversible and specific through competitive binding between the dendrimer and glucose with Con A are presented. Overall, the assay is shown to have potential over the traditional dextran-based assay because it has a larger dynamic response to physiological glucose concentrations, incorporates longer wavelength dyes that improve signal penetration through dermal tissue, and provides an internal reference in the form of a nonreactive fluorescent label.

Microcapsule biosensors using competitive binding resonance energy transfer assays based on apoenzymes

This paper reports the first demonstration of a fluorescence resonance energy transfer based glucose sensor, wherein a competitive binding (CB) assay is encapsulated into polyelectrolyte microcapsules. The work supports the concept that microcapsules are superior to hydrogel systems or other matrixes for competitive-binding-based system, as they provide free movement of the sensing elements within the capsule interior while constant total sensing assay concentration is maintained. The transduction approach employed in these preliminary experiments is also a novel CB system based on a model apoenzyme, apo-glucose oxidase (AG), which is highly specific to beta-d-glucose, as the model target-binding protein. The glucose sensitivity of the fluorescein isothiocyanate (FITC)-dextran and tetramethylrhodamine isothiocyanate-AG encapsulated in microcapsules showed 5 times greater specificity for beta-D-glucose over other sugars, with sensitivity (change in intensity ratio) in the range of 2-6%/mM. It was observed that the sensitivity and range of the response can be tailored by controlling the assay concentration using different FITC-dextran molecular weight and total capsule concentration. The findings support the concepts of using microcapsules to encapsulate CB assays for reversible and stable sensors and the use of apoenzymes as specific molecular recognition elements in CB assays. Further, characterization results for microcapsule glucose sensors demonstrate their suitability for monitoring physiological glucose levels.

Alignment using genetic programming with causal trees for identification of protein functions

A hybrid evolutionary model is used to propose a hierarchical homology of protein sequences to identify protein functions systematically. The proposed model offers considerable potentials, considering the inconsistency of existing methods for predicting novel proteins. Because some novel proteins might align without meaningful conserved domains, maximizing the score of sequence alignment is not the best criterion for predicting protein functions. This work presents a decision model that can minimize the cost of making a decision for predicting protein functions using the hierarchical homologies. Particularly, the model has three characteristics: (i) it is a hybrid evolutionary model with multiple fitness functions that uses genetic programming to predict protein functions on a distantly related protein family, (ii) it incorporates modified robust point matching to accurately compare all feature points using the moment invariant and thin-plate spline theorems, and (iii) the hierarchical homologies holding up a novel protein sequence in the form of a causal tree can effectively demonstrate the relationship between proteins. This work describes the comparisons of nucleocapsid proteins from the putative polyprotein SARS virus and other coronaviruses in other hosts using the model.

Real-Time Assessment of Spatial and Temporal Coupled Catalysis within Polyelectrolyte Microcapsules Containing Coimmobilized Glucose Oxidase and Peroxidase

The encapsulation of biological enzymes within polyelectrolyte microcapsules is an important step toward microscale devices for processing and analytical applications, one which could be applied to the realization of minimally invasive sensing technology. In this work, the encapsulation and functional characterization of a bienzymatic coupled catalytic system within polyelectrolyte microcapsules is described. The two components, glucose oxidase (GOx) and horseradish peroxidase (HRP), were coprecipitated with calcium carbonate microspheres, followed by layer-by-layer assembly to form ultrathin polymer film coatings that act as capsule walls after removal of the sacrificial carbonate cores. Encapsulated concentrations of GOx and HRP were determined to be 19.7 +/- 1.0 and 29.4 +/- 3.6 mg/mL, respectively. An 85% decrease in the rate of glucose consumption relative to GOx and HRP in free solution was observed, which is attributed to substrate diffusion limitations. To further understand the temporal and spatial dynamics of the two-step reaction, a technique for monitoring microscale glucose consumption was developed using confocal imaging techniques. Time-based acquisition of capsule/Amplex Red suspensions was performed, from which it was observed that the high concentration of enzyme immobilized within the capsule walls resulted in a greater rate and quantity of glucose consumption at the capsule periphery when compared to glucose consumption within the capsule interior. These findings demonstrate the function of a bienzymatic catalytic system within the controlled environment of polyelectrolyte microspheres and a novel approach to analysis of the internal reactions using confocal imaging that will allow direct comparison with reaction-diffusion modeling and further explorations to optimize the distribution and activity of the encapsulated species.

Risk factors for symptomatic gallstones in patients with liver cirrhosis: a case-control study

OBJECTIVES

Liver cirrhosis is a well-documented risk factor for the formation of gallstones. In cirrhotic patients, gallstones are almost always "silent," and surgery is rarely required. When indicated (symptoms or complications), cholecystectomy implies a high morbidity risk in these patients, especially in the advanced stages of cirrhosis. The aim of this study was to estimate the risk factors for symptom development in cirrhotic patients with gallstones to identify the subgroup of patients at risk of undergoing surgery.

METHODS

A total of 140 patients with liver cirrhosis and gallstones were studied: 97 with asymptomatic and 43 with symptomatic gallstone disease. The risk factors for gallstone formation (age, gender, family history, parity, obesity, diabetes mellitus, hyperlipoproteinemia) and the characteristics of liver cirrhosis (etiology, duration, Child class, hypersplenism), gallstones (duration, number, size), and gallbladder (size, wall thickness) were assessed in all patients. In 12 patients (four symptomatic, eight asymptomatic), gallbladder emptying was also evaluated by ultrasound. The association of asymptomatic and symptomatic gallstones with all these parameters was statistically evaluated by Student's t, Mann-Whitney, and chi(2) tests, as well as by means of multiple logistic regression. The causal relationship between these characteristics and gallstone symptoms was also examined by means of the KDD (knowledge discovery from databases) method, with an algorithm for learning Bayesian networks.

RESULTS

Advanced age, female gender, viral etiology of cirrhosis, family history of gallstones, and duration of gallstone disease were significantly associated with symptomatic gallstone disease. The number or size of gallstones and the size or emptying of the gallbladder did not differ in symptomatic versus asymptomatic patients. Male gender and alcoholic cirrhosis were inversely correlated with symptom presence. In the multivariate analysis, family history (p = 0.0098) and advanced age (p = 0.0422) were positively correlated and male gender (p = 0.0049) and alcoholic etiology of cirrhosis (p = 0.0116) negatively correlated with symptom presence. These relationships (except for age) were also evidenced by the KDD method.

CONCLUSIONS

The risk of gallstones becoming symptomatic is significantly lower in men and in alcoholic cirrhosis. In cirrhotic women, and especially in the presence of a positive family history and of advanced age, the risk of developing symptoms and undergoing surgery was significantly greater.

Evolutionary computing for knowledge discovery in medical diagnosis

One of the major challenges in medical domain is the extraction of comprehensible knowledge from medical diagnosis data. In this paper, a two-phase hybrid evolutionary classification technique is proposed to extract classification rules that can be used in clinical practice for better understanding and prevention of unwanted medical events. In the first phase, a hybrid evolutionary algorithm (EA) is utilized to confine the search space by evolving a pool of good candidate rules, e.g. genetic programming (GP) is applied to evolve nominal attributes for free structured rules and genetic algorithm (GA) is used to optimize the numeric attributes for concise classification rules without the need of discretization. These candidate rules are then used in the second phase to optimize the order and number of rules in the evolution for forming accurate and comprehensible rule sets. The proposed evolutionary classifier (EvoC) is validated upon hepatitis and breast cancer datasets obtained from the UCI machine-learning repository. Simulation results show that the evolutionary classifier produces comprehensible rules and good classification accuracy for the medical datasets. Results obtained from t-tests further justify its robustness and invariance to random partition of datasets.