Multitask joint learning with graph autoencoders for predicting potential MiRNA-drug associations

The occurrence of many diseases is associated with miRNA abnormalities. Predicting potential drug-miRNA associations is of great importance for both disease treatment and new drug discovery. Most computation-based approaches learn one task at a time, ignoring the information contained in other tasks in the same domain. Multitask learning can effectively enhance the prediction performance of a single task by extending the valid information of related tasks. In this paper, we presented a multitask joint learning framework (MTJL) with a graph autoencoder for predicting the associations between drugs and miRNAs. First, we combined multiple pieces of information to construct a high-quality similarity network of both drugs and miRNAs and then used a graph autoencoder (GAE) to learn their embedding representations separately. Second, to further improve the embedding quality of drugs, we added an auxiliary task to classify drugs using the learned representations. Finally, the embedding representations of drugs and miRNAs were linearly transformed to obtain the predictive association scores between them. A comparison with other state-of-the-art models shows that MTJL has the best prediction performance, and ablation experiments show that the auxiliary task can enhance the embedding quality and improve the robustness of the model. In addition, we show that MTJL has high utility in predicting potential associations between drugs and miRNAs by conducting two case studies.

Amphetamine-type stimulants (ATS) drug classification using shallow one-dimensional convolutional neural network

Amphetamine-type stimulants (ATS) drug analysis and identification are challenging and critical nowadays with the emergence production of new synthetic ATS drugs with sophisticated design compounds. In the present study, we proposed a one-dimensional convolutional neural network (1DCNN) model to perform ATS drug classification as an alternative method. We investigate as well as explore the classification behavior of 1DCNN with the utilization of the existing novel 3D molecular descriptors as ATS drugs representation to become the model input. The proposed 1DCNN model is composed of one convolutional layer to reduce the model complexity. Besides, pooling operation that is a standard part of traditional CNN is not applied in this architecture to have more features in the classification phase. The dropout regularization technique is employed to improve model generalization. Experiments were conducted to find the optimal values for three dominant hyper-parameters of the 1DCNN model which are the filter size, transfer function, and batch size. Our findings found that kernel size 11, exponential linear unit (ELU) transfer function and batch size 32 are optimal for the 1DCNN model. A comparison with several machine learning classifiers has shown that our proposed 1DCNN has achieved comparable performance with the Random Forest classifier and competitive performance with the others.

Visual Comparison of Guidelines: Method and Application to Potentially Inappropriate Medication Lists

Therapeutic guidelines developed by experts are essential tools for improving therapy and drug prescription. Several guidelines often exist that target the same patient, from different organizations and countries. The case of lists for the detection of potentially inappropriate medications (PIMs) is an example which illustrates how these guidelines can be varied and multiple. In order to have an overview to the divergences and similarities between different lists of PIMs, we propose a visual method to compare PIMs lists, based on set visualization, and we apply it to 5 guidelines.

Classification of the present pharmaceutical agents based on the possible effective mechanism on the COVID-19 infection

OBJECTIVES

There are several types of research on the COVID-19 disease which have been conducting. It seems that prevailing over the pandemic would be achieved only by mastering over the virus pathophysiology. We tried to categorize the massive amount of available information for useful interpretation.

EVIDENCE ACQUISITION

We searched databases with different keywords and search strategies that focus on virulence and pathophysiology of COVID-19. The present review has aimed to gather and categorize all implemented drugs based on the susceptible virulence mechanisms, and the pathophysiological events in the host cells, discussing and suggesting treatments.

RESULTS

As a result, the COVID-19 lifecycle were categorized as following steps: "Host Cell Attachment" which is mainly conducted with ACE2 receptors and TMPRSS2 from the host cell and Spike (S) protein, "Endocytosis Pathway" which is performed mainly by clathrin-mediated endocytosis, and "Viral Replication" which contains translation and replication of RNA viral genome. The virus pathogenicity is continued by "Inflammatory Reactions" which mainly caused moderate to severe COVID-19 disease. Besides, the possible effective therapeutics' mechanism and the pharmaceutical agents that had at least one experience as a preclinical or clinical study on COVID-19 were clearly defined.

CONCLUSION

The treatment protocol would be occasional based on the stage of the infection and the patient situation. The cocktail of medicines, which could affect almost all mentioned stages of COVID-19 disease, might be vital for patients with severe phenomena. The classification of the possible mechanism of medicines based on COVID-19 pathogenicity.

On subclasses of opioid analgesics

BACKGROUND

The history of discovery of analgesic drugs has followed a trajectory from original serendipitous discovery of plant-derived substances to laboratory creation of customized molecules that are intentionally designed to interact with specific receptors of neurotransmitters involved in either the transmission of the pain signal or the attenuation of such a signal. The drugs most recently developed have been designed to provide incremental greater separation between pain relief and adverse effects. The result has been drugs that have individualized pharmacodynamic and pharmacokinetic characteristics that represent specific advances in basic science and translate into unique clinical profiles. Several of the drugs include non-opioid components. They retain some of the features of opioids, but have distinct clinical characteristics that differentiate them from traditional opioids. Thus they defy simple classification as opioids.

SCOPE

A summary is provided of the development of the modern view of multi-mechanistic pain and its treatment using analgesics that have multi-mechanisms of action (consisting of both opioid and non-opioid components). Descriptions of examples of such current analgesics and of those that have pharmacokinetic characteristics that result in atypical opioid clinical profiles are given.

FINDINGS

By serendipity or design, several current strong analgesics have opioid components of action, but have an additional non-opioid mechanism of action or some pharmacokinetic feature that gives them an atypical opioid clinical profile and renders them not easily classified as classical opioids.

CONCLUSION

An appreciation that there are now opioid analgesics that differentiate from classical opioids in ways that defy their simplistic classification as opioids suggests that recognition of subclasses of opioid analgesics would be more accurate scientifically and would be more informative for healthcare providers and regulators. This would likely lead to positive outcomes for the clinical use and regulatory control of the current drugs, and provide direction/strategy for the discovery of new drugs.

How do researchers categorize drugs, and how do drug users categorize them?

This paper considers drug classifications and terms widely used in US survey research, and compares these to classifications and terms used by drug users. We begin with a critical review of drug classification systems, including those oriented to public policy and health services as well as survey research. We then consider the results of a pile sort exercise we conducted with 76 respondents within a mixed method study of Southeast Asian American adolescent and young adult drug users in urban Northern California, USA. We included the pile sort to clarify how respondents handled specific terms which we understood to be related to Ecstasy and methamphetamines. Results of the pile sort were analyzed using graphic layout algorithms as well as content analysis of pile labels. Similar to the national surveys, our respondents consistently differentiated Ecstasy terms from methamphetamine terms. We found high agreement between some specific local terms (thizz, crystal) and popular drug terms, while other terms thought to be mainstream (crank, speed) were reported as unknown by many respondents. In labeling piles, respondents created taxonomies based on consumption method (in particular, pill) as well as the social contexts of use. We conclude by proposing that divergences between drug terms utilized in survey research and those used by drug users may reflect two opposing tendencies: the tendency of survey researchers to utilize standardized language that constructs persons and experiences as relatively homogeneous, varying only within measurable degrees, and the tendency of drug users to utilize specialized language (argot) that reflects their understandings of their experiences as hybrid and diverse. The findings problematize the validity of drug terms and categories used in survey research.