Determination of the lactose content in low-lactose milk using Fourier-transform infrared spectroscopy (FTIR) and convolutional neural network.

Demand for low lactose milk and milk products has been increasing worldwide due to the high number of people with lactose intolerance. These low lactose dairy foods require fast, low-cost and efficient methods for sugar quantification. However, available methods do not meet all these requirements. In this work, we propose the association of FTIR (Fourier Transform Infrared) spectroscopy with artificial intelligence to identify and quantify residual lactose and other sugars in milk. Convolutional neural networks (CNN) were built from the infrared spectra without preprocessing the data using hyperparameter adjustment and saliency map. For the quantitative prediction of the sugars in milk, a regression model was proposed, while for the qualitative assessment, a classification model was used. Raw, pasteurized and ultra-high temperature (UHT) milk was added with lactose, glucose, and galactose in six concentrations (0.1-7.0 mg mL-1) and, in total, 432 samples were submitted to convolutional neural network. Accuracy, precision, sensitivity, specificity, root mean square error, mean square error, mean absolute error, and coefficient of determination (R2) were used as evaluation parameters. The algorithms indicated a predictive capacity (accuracy) above 95% for classification, and R2 of 81%, 86%, and 92% for respectively, lactose, glucose, and galactose quantification. Our results showed that the association of FTIR spectra with artificial intelligence tools, such as CNN, is an efficient, quick, and low-cost methodology for quantifying lactose and other sugars in milk.

A machine learning proposal method to detect milk tainted with cheese whey.

Cheese whey addition to milk is a type of fraud with high prevalence and severe economic effects, resulting in low yield for dairy products, nutritional reduction of milk and milk-derived products, and even some safety concerns. Nevertheless, methods to detect fraudulent addition of cheese whey to milk are expensive and time consuming, and are thus ineffective as screening methods. The Fourier-transform infrared (FTIR) spectroscopy technique is a promising alternative to identify this type of fraud because a large number of data are generated, and useful information might be extracted to be used by machine learning models. The objective of this work was to evaluate the use of FTIR with machine learning methods, such as classification tree and multilayer perceptron neural networks to detect the addition of cheese whey to milk. A total of 520 samples of raw milk were added with cheese whey in concentrations of 1, 2, 5, 10, 15, 20, 25, and 30%; and 65 samples were used as control. The samples were stored at 7, 20, and 30°C for 0, 24, 48, 72, and 168 h, and analyzed using FTIR equipment. Complementary results of 520 samples of authentic raw milk were used. Selected components (fat, protein, casein, lactose, total solids, and solids nonfat) and freezing point (°C) were predicted using FTIR and then used as input features for the machine learning algorithms. Performance metrics included accuracy as high as 96.2% for CART (classification and regression trees) and 97.8% for multilayer perceptron neural networks, with precision, sensitivity, and specificity above 95% for both methods. The use of milk composition and freezing point predicted using FTIR, associated with machine learning techniques, was highly efficient to differentiate authentic milk from samples added with cheese whey. The results indicate that this is a potential method to be used as a high-performance screening process to detected milk adulterated with cheese whey in milk quality laboratories.

A System for Structuring, Storage and Georeferenciation of Dengue Vector Surveillance Data.

Dengue is a main public health issue around the world and is an epidemic in Brazil. As part of the Brazilian national program to fight the disease, every municipality has a Zoonosis Control Center responsible for health and case surveillance, among other actions. The fieldwork includes routine visiting of houses and strategic sites (e.g. industries and vacant lands), water sampling, container elimination, and larvicide administration. However, the field data are gathered and summarized by hand. In this work, our goal is to ease the collection and visualization of field data to support decision-making. We have developed a mobile system to collect and georeference field data which could then be used to build geospatial and geo-temporal visualizations of indices such as House, Container, and Breteau1 indices. This solution could enhance entomological surveillance and leverage action planning and evaluation.

In Silico Laboratory Experiments Using Statistical Model Checking: A New Model of the Palytoxin-Induced Pump Channel as Case Study.

Studying biological systems is a difficult but important task. Traditional methods include laboratory experimentation and computer simulations. However, often researchers need to explore important but potentially rare events that are not easily observed or simulated. We use UPPAAL-SMC, a formal verification tool to support a methodology that allows us to model biological systems, specify events and conditions that we want to analyze, and to explore system executions using controlled simulations. We also describe an efficient way to reproduce laboratory experiments in silico. Unlike traditional simulations, we are able to guide the experiment to explore special events and conditions by expressing these conditions in temporal logic formulas. We have applied this methodology to create a more detailed model of Palytoxin-induced Na +/K + pump channels than was previously possible. Moreover, we have reproduced experimental protocols and their associated electrophysiological recordings, which has not been done in previous works. As a consequence, we have been able to propose a new diprotomeric model for the PTX-pump complex and study its behaviour. The use of our methodology has enabled us to reduce the effort and time to perform this research. It can be used to model and analyze other complex biological systems, potentially increasing the productivity of such studies.

On the utilization of deep and ensemble learning to detect milk adulteration.

BACKGROUND: Fraudulent milk adulteration is a dangerous practice in the dairy industry that is harmful to consumers since milk is one of the most consumed food products. Milk quality can be assessed by Fourier Transformed Infrared Spectroscopy (FTIR), a simple and fast method for obtaining its compositional information. The spectral data produced by this technique can be explored using machine learning methods, such as neural networks and decision trees, in order to create models that represent the characteristics of pure and adulterated milk samples. RESULTS: Thousands of milk samples were collected, some of them were manually adulterated with five different substances and subjected to infrared spectroscopy. This technique produced spectral data from the milk samples composition, which were used for training different machine learning algorithms, such as deep and ensemble decision tree learners. The proposed method is used to predict the presence of adulterants in a binary classification problem and also the specific assessment of which of five adulterants was found through multiclass classification. In deep learning, we propose a Convolutional Neural Network architecture that needs no preprocessing on spectral data. Classifiers evaluated show promising results, with classification accuracies up to 98.76%, outperforming commonly used classical learning methods. CONCLUSIONS: The proposed methodology uses machine learning techniques on milk spectral data. It is able to predict common adulterations that occur in the dairy industry. Both deep and ensemble tree learners were evaluated considering binary and multiclass classifications and the results were compared. The proposed neural network architecture is able to outperform the composition recognition made by the FTIR equipment and by commonly used methods in the dairy industry.

Hypothesized diprotomeric enzyme complex supported by stochastic modelling of palytoxin-induced Na/K pump channels.

The sodium-potassium pump (Na+/K+ pump) is crucial for cell physiology. Despite great advances in the understanding of this ionic pumping system, its mechanism is not completely understood. We propose the use of a statistical model checker to investigate palytoxin (PTX)-induced Na+/K+ pump channels. We modelled a system of reactions representing transitions between the conformational substates of the channel with parameters, concentrations of the substates and reaction rates extracted from simulations reported in the literature, based on electrophysiological recordings in a whole-cell configuration. The model was implemented using the UPPAAL-SMC platform. Comparing simulations and probabilistic queries from stochastic system semantics with experimental data, it was possible to propose additional reactions to reproduce the single-channel dynamic. The probabilistic analyses and simulations suggest that the PTX-induced Na+/K+ pump channel functions as a diprotomeric complex in which protein-protein interactions increase the affinity of the Na+/K+ pump for PTX.

An innovative electronic health records system for rare and complex diseases.

BACKGROUND: There exists a large number of rare and complex diseases that are neglected due to the difficulty in diagnosis and treatment. Being rare, they normally do not justify the costs of developing an especialized Electronic Health Record (EHR) system to assist doctors and patients of these diseases. In this work we propose the use of Computer applications known as Laboratory Information Management Systems (LIMS) to address this issue. RESULTS: In this work we describe a fully customizable EHR system that uses a workflow based LIMS with an easy to adapt interface for data collection and retrieval. This system can easily be customized to manage different types of medical data. The customization for a new disease can be done in a few hours with the help of a specialist. CONCLUSION: We have used the proposed system to manage data from patients of three complex diseases: neuromyelitis optica, paracoccidioidomycosis and adrenoleukodistrofy. These diseases have very different symptoms, exams, diagnostics and treatments, but the FluxMED system is able to manage these data in a highly specialized manner without any modifications to its code.

FluxCTTX: A LIMS-based tool for management and analysis of cytotoxicity assays data.

BACKGROUND: Cytotoxicity assays have been used by researchers to screen for cytotoxicity in compound libraries. Researchers can either look for cytotoxic compounds or screen "hits" from initial high-throughput drug screens for unwanted cytotoxic effects before investing in their development as a pharmaceutical. These assays may be used as an alternative to animal experimentation and are becoming increasingly important in modern laboratories. However, the execution of these assays in large scale and different laboratories requires, among other things, the management of protocols, reagents, cell lines used as well as the data produced, which can be a challenge. The management of all this information is greatly improved by the utilization of computational tools to save time and guarantee quality. However, a tool that performs this task designed specifically for cytotoxicity assays is not yet available. RESULTS: In this work, we have used a workflow based LIMS -- the Flux system -- and the Together Workflow Editor as a framework to develop FluxCTTX, a tool for management of data from cytotoxicity assays performed at different laboratories. The main work is the development of a workflow, which represents all stages of the assay and has been developed and uploaded in Flux. This workflow models the activities of cytotoxicity assays performed as described in the OECD 129 Guidance Document. CONCLUSIONS: FluxCTTX presents a solution for the management of the data produced by cytotoxicity assays performed at Interlaboratory comparisons. Its adoption will contribute to guarantee the quality of activities in the process of cytotoxicity tests and enforce the use of Good Laboratory Practices (GLP). Furthermore, the workflow developed is complete and can be adapted to other contexts and different tests for management of other types of data.

Probabilistic model checking analysis of palytoxin effects on cell energy reactions of the Na+/K+-ATPase.

Probabilistic model checking (PMC) is a technique used for the specification and analysis of complex systems. It can be applied directly to biological systems which present these characteristics, including cell transport systems. These systems are structures responsible for exchanging ions through the plasma membrane. Their correct behavior is essential for animal cells, since changes on those are responsible for diseases. In this work, PMC is used to model and analyze the effects of the palytoxin toxin (PTX) interactions with one of these systems. Our model suggests that ATP could inhibit PTX action. Therefore, individuals with ATP deficiencies, such as in brain disorders, may be more susceptible to the toxin. We have also used heat maps to enhance the kinetic model, which is used to describe the system reactions. The map reveals unexpected situations, such as a frequent reaction between unlikely pump states, and hot spots such as likely states and reactions. This type of analysis provides a better understanding on how transmembrane ionic transport systems behave and may lead to the discovery and development of new drugs to treat diseases associated to their incorrect behavior.

NMO-DBr: the Brazilian Neuromyelitis Optica Database System.

OBJECTIVE: To present the Brazilian Neuromyelitis Optica Database System (NMO-DBr), a database system which collects, stores, retrieves, and analyzes information from patients with NMO and NMO-related disorders. METHOD: NMO-DBr uses Flux, a LIMS (Laboratory Information Management Systems) for data management. We used information from medical records of patients with NMO spectrum disorders, and NMO variants, the latter defined by the presence of neurological symptoms associated with typical lesions on brain magnetic resonance imaging (MRI) or aquaporin-4 antibody seropositivity. RESULTS: NMO-DBr contains data related to patient's identification, symptoms, associated conditions, index events, recurrences, family history, visual and spinal cord evaluation, disability, cerebrospinal fluid and blood tests, MRI, optic coherence tomography, diagnosis and treatment. It guarantees confidentiality, performs cross-checking and statistical analysis. CONCLUSION: NMO-DBr is a tool which guides professionals to take the history, record and analyze information making medical practice more consistent and improving research in the area.

NMO-DBr: the Brazilian Neuromyelitis Optica Database System / NMO-DBr: o banco de dados brasileiro em neuromielite óptica

OBJECTIVE: To present the Brazilian Neuromyelitis Optica Database System (NMO-DBr), a database system which collects, stores, retrieves, and analyzes information from patients with NMO and NMO-related disorders. METHOD: NMO-DBr uses Flux, a LIMS (Laboratory Information Management Systems) for data management. We used information from medical records of patients with NMO spectrum disorders, and NMO variants, the latter defined by the presence of neurological symptoms associated with typical lesions on brain magnetic resonance imaging (MRI) or aquaporin-4 antibody seropositivity. RESULTS: NMO-DBr contains data related to patient's identification, symptoms, associated conditions, index events, recurrences, family history, visual and spinal cord evaluation, disability, cerebrospinal fluid and blood tests, MRI, optic coherence tomography, diagnosis and treatment. It guarantees confidentiality, performs cross-checking and statistical analysis. CONCLUSION: NMO-DBr is a tool which guides professionals to take the history, record and analyze information making medical practice more consistent and improving research in the area.

OBJETIVO: Apresentar o Brazilian Neuromyelitis Optica Database System (NMO-DBr), um sistema de banco de dados que coleta, arquiva, recupera e analisa informações de pacientes com neuromielite óptica (NMO) e doenças relacionadas. MÉTODO: NMO-DBr usa o sistema Flux, um LIMS (Laboratory Information Management Systems) para gerenciamento de informações. As informações foram colhidas dos prontuários de pacientes com espectro de NMO e variantes de NMO, estas últimas definidas por quadro neurológico associado a lesões encefálicas típicas à imagem pela ressonância magnética (IRM) ou à soropositividade do anticorpo anti-aquaporina-4. RESULTADOS: NMO-DBr contém dados relativos a identificação, sintomas, condições associadas, eventos índices, recorrências, história familiar, avaliação visual e da medula, incapacidade, exames do líquor e de sangue, IRM, tomografia de coerência óptica (OCT), diagnóstico e tratamento. O sistema assegura confidencialidade, cruza dados e faz análises estatísticas. CONCLUSÃO: NMO-DBr é uma ferramenta que possibilita a prática médica mais consistente e promove a pesquisa na área.

Can the vector space model be used to identify biological entity activities?

BACKGROUND: Biological systems are commonly described as networks of entity interactions. Some interactions are already known and integrate the current knowledge in life sciences. Others remain unknown for long periods of time and are frequently discovered by chance. In this work we present a model to predict these unknown interactions from a textual collection using the vector space model (VSM), a well known and established information retrieval model. We have extended the VSM ability to retrieve information using a transitive closure approach. Our objective is to use the VSM to identify the known interactions from the literature and construct a network. Based on interactions established in the network our model applies the transitive closure in order to predict and rank new interactions. RESULTS: We have tested and validated our model using a collection of patent claims issued from 1976 to 2005. From 266,528 possible interactions in our network, the model identified 1,027 known interactions and predicted 3,195 new interactions. Iterating the model according to patent issue dates, interactions found in a given past year were often confirmed by patent claims not in the collection and issued in more recent years. Most confirmation patent claims were found at the top 100 new interactions obtained from each subnetwork. We have also found papers on the Web which confirm new inferred interactions. For instance, the best new interaction inferred by our model relates the interaction between the adrenaline neurotransmitter and the androgen receptor gene. We have found a paper that reports the partial dependence of the antiapoptotic effect of adrenaline on androgen receptor. CONCLUSIONS: The VSM extended with a transitive closure approach provides a good way to identify biological interactions from textual collections. Specifically for the context of literature-based discovery, the extended VSM contributes to identify and rank relevant new interactions even if these interactions occur in only a few documents in the collection. Consequently, we have developed an efficient method for extracting and restricting the best potential results to consider as new advances in life sciences, even when indications of these results are not easily observed from a mass of documents.

Modeling and analysis of cell membrane systems with probabilistic model checking.

BACKGROUND: Recently there has been a growing interest in the application of Probabilistic Model Checking (PMC) for the formal specification of biological systems. PMC is able to exhaustively explore all states of a stochastic model and can provide valuable insight into its behavior which are more difficult to see using only traditional methods for system analysis such as deterministic and stochastic simulation. In this work we propose a stochastic modeling for the description and analysis of sodium-potassium exchange pump. The sodium-potassium pump is a membrane transport system presents in all animal cell and capable of moving sodium and potassium ions against their concentration gradient. RESULTS: We present a quantitative formal specification of the pump mechanism in the PRISM language, taking into consideration a discrete chemistry approach and the Law of Mass Action aspects. We also present an analysis of the system using quantitative properties in order to verify the pump reversibility and understand the pump behavior using trend labels for the transition rates of the pump reactions. CONCLUSIONS: Probabilistic model checking can be used along with other well established approaches such as simulation and differential equations to better understand pump behavior. Using PMC we can determine if specific events happen such as the potassium outside the cell ends in all model traces. We can also have a more detailed perspective on its behavior such as determining its reversibility and why its normal operation becomes slow over time. This knowledge can be used to direct experimental research and make it more efficient, leading to faster and more accurate scientific discoveries.

Production of full-length cDNA sequences by sequencing and analysis of expressed sequence tags from Schistosoma mansoni

The number of sequences generated by genome projects has increased exponentially, but gene characterization has not followed at the same rate. Sequencing and analysis of full-length cDNAs is an important step in gene characterization that has been used nowadays by several research groups. In this work, we have selected Schistosoma mansoni clones for full-length sequencing, using an algorithm that investigates the presence of the initial methionine in the parasite sequence based on the positions of alignment start between two sequences. BLAST searches to produce such alignments have been performed using parasite expressed sequence tags produced by Minas Gerais Genome Network against sequences from the database Eukaryotic Cluster of Orthologous Groups (KOG). This procedure has allowed the selection of clones representing 398 proteins which have not been deposited as S. mansoni complete CDS in any public database. Dedicated sequencing of 96 of such clones with reads from both 5' and 3' ends has been performed. These reads have been assembled using PHRAP, resulting in the production of 33 full-length sequences that represent novel S. mansoni proteins. These results shall contribute to construct a more complete view of the biology of this important parasite.

Production of full-length cDNA sequences by sequencing and analysis of expressed sequence tags from Schistosoma mansoni.

The number of sequences generated by genome projects has increased exponentially, but gene characterization has not followed at the same rate. Sequencing and analysis of full-length cDNAs is an important step in gene characterization that has been used nowadays by several research groups. In this work, we have selected Schistosoma mansoni clones for full-length sequencing, using an algorithm that investigates the presence of the initial methionine in the parasite sequence based on the positions of alignment start between two sequences. BLAST searches to produce such alignments have been performed using parasite expressed sequence tags produced by Minas Gerais Genome Network against sequences from the database Eukaryotic Cluster of Orthologous Groups (KOG). This procedure has allowed the selection of clones representing 398 proteins which have not been deposited as S. mansoni complete CDS in any public database. Dedicated sequencing of 96 of such clones with reads from both 5' and 3' ends has been performed. These reads have been assembled using PHRAP, resulting in the production of 33 full-length sequences that represent novel S. mansoni proteins. These results shall contribute to construct a more complete view of the biology of this important parasite.

Efficient secondary database driven annotation using model organism sequences.

The use of sequences from specific organisms for annotation requires that it does not represent great loss of information and that the sequences available suffice for annotation. In order to investigate whether or not sequences from model organisms may suffice for annotation of sequences from the trematode Schistosoma mansoni, we performed local BLAST searches of S. mansoni sequences against other organisms sequences present in the NCBI database nr. Results have been inserted into a relational database and hits to sequences from three model organisms, Caenorhabditis elegans, Drosophila melanogaster and Homo sapiens have been computed and compared to hits to sequences from other organisms present in nr; score values of each alignment have also been registered. Our observations have shown that a large fraction of orthologous proteins exists in the set of sequences from the three model organisms selected, and therefore a similar fraction of transcripts can be annotated when using either nr or model organism datasets. Moreover, hits to model organisms' sequences are largely as informative as nr. Results suggest that model organisms provide a reliable set of sequences to use as a reference database for S. mansoni sequence annotation, showing the clear possibility of using a restricted dataset of expected better quality for functional annotation and therefore supporting secondary database driven annotation approaches.