In this communication we carried a research in order to establish mathematical expressions that relate the regrowth age with the sugar content (glucose, fructose and sucrose) of the whole plant and its fractions. The experiment was developed, following a randomized block design, with 5 treatments (60, 90, 120, 150 and 180 d). In the first part of this communication the content of Nitrogen (N) in the integral plant, leaves and stems were evaluated. Reporting that Glu, Frut and Suc decreased with the highest results at 60 days and quadratic and cubic equations with R2 higher than 0.90 were adjusted.

In this work we report a Machine Learning study of a dataset involving the α-amylase inhibitors. The prediction of α-amylase inhibitory activity as anti-diabetic is carried out using LDA and classification trees (CT). A large data set of 640 compounds for α-amylase was selected to developing the ML models. In the case of CT-J48 have the better classification model performances with values above 80- 90% for the training and prediction sets, correspondingly. The best model shows an accuracy higher than 95% for training set; the model was also validated using 10-fold cross-validation procedure and through a test set achieving accuracies values of 85.32% and 86.80%, correspondingly. The full paper was published in Chem Biol Drug Des. 2019;00:1–8. DOI: 10.1111/cbdd.13518

BACKGROUND：This retrospective study aimed to investigate pregnancy outcomes when transferred three different sources embryos during early rescue ICSI.

MATERIAL AND METHOD: A total of 805 infertility cases were included. On the third day after insemination, 615 cases transferred with one to three embryos. The pregnant rate, abortion rate and live birth rate were compared in three groups.

RESULTS: ①Pregnant group with younger female age, less starting gonadotropin dose, fertilization and embryonic score were better. ②Pregnant rate, early abortion rate and ectopic pregnancy rate were the lowest in only ICSI sources embryo. Multiple pregnancy rate and the birth rate of low weight babies at term was lower in the transferred both IVF and ICSI sources embryo group. ③The sex ratio of the newborn in the three groups were different significantly, IVF with more girl, ICSI with more boy, and IVF/ICSI group was the most balanced. ④Both cleavage rate and good quality embryo rate were the predictors of clinical pregnancy and live birth.

CONCLUSIONS: Using the early rescue ICSI based on short-term insemination was helpful to increase the utilization rate of the embryo, and improve the clinical pregnancy outcome. While more researches are needed on the safety of offspring.

Today, studies are performed from a dataset spanning multiple preclinical assays and different experimental conditions for sarcomas. PTML is a tool that combines Machine Learning (ML) algorithms and Perturbation Theory (PT) principles. With PTML, ML techniques can be used to predict antisarcoma compounds. At the same time, different PT techniques can be applied. One of the most widely used ML techniques is the neural network which showed high accuracy for both training and model validation. It is important to emphasize that the production of the most optimal model would save resources in the pharmaceutical industries. In a recent paper Cabrera et al. reported a new model for prediction of anti-sarcoma compounds. The model is very interesting because it can predict the biological activity vs multiple proteins etc. The authors also explored multiple molecular descriptors of drugs as well as many assay conditions like protein target, cell line, etc. There are some suggestions we can make to improve future versions of this paper. For instance, the authors could calculate also sequence descriptor of target proteins to predict the results for new mutants. On my opinion, it could be very interesting developing a user friendly software for use of non expert medicinal chemists. This software could be a desktop or online server application increasing the use of the model worldwide. Another interesting step could be the fusion of the present pre-clinical data with clinical data including variables of patients or population groups. In all case, the paper is very interesting an opens new gates to the authors for future works including new features to the design of antisarcoma compounds.

Patient who suffer for Cystic Fibrosis (CF) have a lot of problems in their lives. For avoid that this patient have to follow a specific treatment which, in some moment, can be difficult. Because of that, nowadays is using Machine Learning (ML) techniques to improve the care of the patient with CF. Taking advantage of high-quality data including many variables and longitudinal biomarkers of thousands of patients over long periods of time, ML technique is helping in the clinical work, as it do a personalized approach to patient management with this disease. ML-based models can guide personalized decisions through individual-level predictions on: whether the patient at hand is likely to have an exacerbation, whether a specific new treatment is likely to be effective for that patient, what is the relative probability of their different competing risk, whether a proactive clinical intervention and clinical visit is needed, etc.

In conclusion, the use of ML-based model helps medical personnel improving the care of these patients. For that, clinicians must follow this order that will help them making different predictions: Risk prediction, prediction the trajectory of the individual patient to understand better the disease, predict competing risk, predict which treatment is more suitable for an individual given their particular characteristics, predictions about the patients´ health and progress including recommendations (for example alternative of diet and exercise or alert the patient and medical personnel when further action or consultation is needed), and finally doing predictions to discover the clinical significance of specific characteristics that were not previously understood to be important.

This paper explained the development of an algorithmic ``AutoPrognosis´´ that used Machine Learning (ML) to automate the process of constructing clinical prognostic model for the Cystic Fibrosis (CF) (but can be applied for any other disease), as well as the risk factors of this disease. For this construction it has been taken into account some rules associated to risk factors, for example that the main predictor of mortality in CF patients is having the Forced Expiratory Volume (FEV₁) biomarker higher than 30%.

This model needs to be updated and re-calibrated annually. Also they conducted an extensive analysis of the performance of AutoPrognosis, and compared to those achieved by the existing guidelines, competing clinical models and other ML algorithms.

The method of how this model works is provided in the next figure. AutoPrognosis uses Bayesian optimization technique in order to improve ML. With this is predefined an accurate diagnostic. This consists of an imputation algorithm, a feature processing algorithm, a classification algorithm and a calibration method where all of these are combined in a single. The final stage of AutoPrognosis is an interpreter module, which uses an associative classifier to explain the learned predictions.

Although nowadays the practice and deployment of this type of model in healthcare research has been limited, using it is a good way to have a previous idea of what is going to happen to a patient that suffers from CF. This is known because after doing this study AutoPrognosis achieves significant gains in terms of the positive predictive values (for example implies a remarkable improvement in terms of the precision of lung transplant referral decisions). Nevertheless, it has some limitations: the need to be extremely validated, need to be valuated by considering post-transplant survival data, and as they did not have access for data on patients who went through a transplant evaluation process or were enrolled in wait list but did not get a transplant within the 3-year analysis horizon, it is rendered direct comparisons with the actually realized clinical policy impossible.

In the medical and pharmaceutical areas, discovering different drugs is too important. This process is long, complex, and depends on numerous factors. To avoid that, Machine Learning (ML) techniques are applied. These techniques are the practice of using algorithms to parse data, learn from it and then make a determination or a prediction about the future state of any new data sets. This practice consists of at least 80% data processing and cleaning and 20% algorithm application. Data types can include images, textual information, biometrics and other information from wearables, assay information, and high-dimensional omics data. In relation with the types of techniques that are used to apply ML can separate into two different. The first one is supervised learning, which is used to develop training models to predict future values of data categories or continuous variables. The second one is unsupervised learning, whereas this one is used for exploratory purposes to develop models that enable clustering of the data in a way that is not specified by the user.

ML presents new opportunities for early target identification and validation that will help us to discover drugs sooner. Apart from being used for that, ML can predict biomarkers, which will be a way to better understand the mechanism of action of a drug and for instance to identify the right drug for the right patients. Moreover, it is used to analysis of digital pathology data in clinical trials. It can also predict disease-specific drug effects.

To sum up, nowadays ML approaches are beginning to be commonly used in the various steps of the discovery and development of drugs by pharmaceutical companies

Deep neural networks are becoming more and more popular due to their revolutionary success in diverse areas, such as computer vision, natural language processing, and speech recognition. However, the decision-making processes of these models are generally not interpretable to users. In various domains, such as healthcare, finance, or law, it is critical to know the reasons behind a decision made by an artificial intelligence system. Therefore, several directions for explaining neural models have recently been explored In this communication, We investigate two major directions for explaining deep neural networks. The first direction consists of feature-based post-hoc explanatory methods, that is, methods that aim to explain an already trained and fixed model (post-hoc), and that provide explanations in terms of input features, such as superpixels for images (feature-based). The second direction consists of self-explanatory neural models that generate medical imaging explanations, that is, models that have a built-in module that generates explanations for the predictions of the model.

The Ecuadorian Amazon region presents a progressive growth of the pig sector, in recent years a way has been sought to obtain a technician production to improve productive and reproductive yields with low levels of environmental pollution and that promote animal welfare at a minimum cost. The objective of this work was to describe the current situation of the deep bedding management system in swine production in the Ecuadorian Amazon region based on the review of scientific information. The present study was exploratory and was based on a bibliographic compilation. The implementation of the deep bedding system in the Ecuadorian Amazon allows to raise pigs with efficient management, in which different adsorbent materials are being used (rice husk, sawdust, shavings and cane bagasse), its implementation allows to reduce the construction costs of facilities, daily cleaning is not carried out, there is a reduction in personnel costs, there is optimization of the consumption of drinking water and cleaning activities, reduction in the emission of odors and liquid and solid waste in tributaries and effluents that produce bad odor, the presence of flies, rodents and other species of animals outside the pig farm. The production of pigs in deep bedding contributes to the improvement of productive indicators, reduces the emission of polluting gases, minimizes the production costs of the system, and guarantees compliance with animal welfare regulations.

Deep neural networks are becoming more and more popular due to their revolutionary success in diverse areas, such as computer vision, natural language processing, and speech recognition. However, the decision-making processes of these models are generally not interpretable to users. In various domains, such as healthcare, finance, or law, it is critical to know the reasons behind a decision made by an artificial intelligence system. Therefore, several directions for explaining neural models have recently been explored.

In this abstract, We investigate two major directions for explaining deep neural networks. The first direction consists of feature-based post-hoc explanatory methods, that is, methods that aim to explain an already trained and fixed model (post-hoc), and that provide explanations in terms of input features, such as superpixels for images (feature-based). The second direction consists of self-explanatory neural models that generate medical imaging explanations, that is, models that have a built-in module that generates explanations for the predictions of the model.