Search

Background Persistent infection with high-risk types Human Papillomavirus could cause diseases including cervical cancers and oropharyngeal cancers. Nonetheless, so far there is no effective pharmacotherapy for treating the infection from high-risk HPV types, and hence it remains to be a severe threat to health of female. Methods In light of drug repositioning strategy, we trained and benchmarked multiple machine learning predictive models so as to predict potential effective antiviral drugs for HPV infection in this work. Based on antiviral-target interaction dataset, we generated high dimension feature set of drug-target interaction pairs and used the dataset to train and construct machine learning predictive models. Results Through optimizing models, measuring models’ predictive performance using 182 pairs of antiviral-target interaction dataset which were all approved by United States Food and Drug Administration, and benchmarking different models’ predictive performance, we identified the optimized Support Vector Machine and K-Nearest Neighbor classifier with high precision score were the best two predictors (0.80 and 0.85 respectively) amongst classifiers of Support Vector Machine, Random forest, Adaboost, Naive Bayes, K-Nearest Neighbors, and Logistic regression classifier. We applied these two predictors together and successfully predicted 58 pairs of antiviral-HPV protein interactions from 846 pairs of antiviral-HPV protein associations. Conclusions Our work provided good drug candidates for anti-HPV drug discovery. So far as we know, we are the first one to conduct such HPV-oriented computational drug repositioning study.

High-throughput sequencing is gaining popularity in clinical diagnoses, but more and more novel gene variants with unknown clinical significance are being found, giving difficulties to interpretations of people’s genetic data, precise disease diagnoses, and the making of therapeutic strategies and decisions. In order to solve these issues, it is of critical importance to figure out ways to analyze and interpret such variants. In this work, BRCA1 gene variants with unknown clinical significance were identified from clinical sequencing data, and then, we developed machine learning models so as to predict the pathogenicity for variants with unknown clinical significance. Through performance benchmarking, we found that the optimized random forest model scored 0.85 in area under receiver operating characteristic curve, which outperformed other models. Finally, we applied the best random forest model to predict the pathogenicity of 6321 BRCA1 variants from both sequencing data and ClinVar database. As a result, we obtained the predictive pathogenic risks of BRCA1 variants of unknown significance.

The difficulty of early diagnosis for ovarian cancer is an important cause of the high mortal rates of ovarian cancer patients. Instead of symptom-based diagnostic methods, modern sequencing technologies enable the access of human’s genetic information via reading DNA/RNA molecules’ nucleotide base sequences. In such way, genes’ mutations and variants could be identified and hence a better clinical diagnosis in molecular level could be expected. However, as sequencing technologies gain more popularity, novel gene variants with unknown clinical significance are found, giving difficulties to interpretations of patients’ genetic data, precise disease diagnoses as well as the making of therapeutic strategies and decisions. In order to solve these issues, it is of critical importance to figure out ways to analyze and interpret such variants. In this work, BRCA1 gene variants with unknown clinical significance were identified from clinical sequencing data, and then we developed machine learning models so as to predict the pathogenicity for variants with unknown clinical significance. Amongst, in performance benchmarking, our optimized random forest model scored 0.85 in area under receiver-operating characteristic curve, which outperformed other models. Finally, we applied the optimized random forest model to predict the pathogenic risks of 7 BRCA1 variants of unknown clinical significances identified from our sequencing data, and 6315 variants of unknown clinical significance in ClinVar database. As a result, our model predicted 4724 benign and 1591 pathogenic variants, which helped the interpretation of these variants of unknown significance and diagnosis.

Pharmaceutical technology transfer is one of the components of pharmaceutical innovation. Currently, a gap exists in pharmaceutical technology transfer from academia to industry. This study aims to develop an objective model to identify valuable pharmaceutical technologies for transferring in order to drive pharmaceutical innovation. We created a support vector machine classifier model using the data of pharmaceutical patents held by universities to predict the licensing outcomes of those patents. We collected data on 369 United States (US) pharmaceutical patents, using 142 licensed patents as the positive samples and 227 unlicensed patents as the negative samples. We also collected the licensing data of the patents, and the distinguished patent features were selected for model training and generation. Upon optimization, the machine learning model was evaluated using different scoring methods. Our support vector machine-based model achieved a fairly good performance of 82.50% in precision and 88.89% in specificity. To the best of our knowledge, our study is the first to apply the machine learning approach to predict the licensing outcomes for pharmaceutical patent valuation and technology transfer. Our work is a good alternative to the current patent valuation methods available in the market, and it could be further developed for practical use in real business contexts.

Today, over 20 million people suffer from Alzheimer's disease (AD) worldwide. AD has become a critical issue to human health, especially in aging societies, and therefore it is a research hotspot in the global scientific community. The technology flow method differs from traditional reviews generating an informative overview of the research and development (RD) landscape in a specific technological area. We need such an updated method to get a general overview of the RD of anti-AD drugs in light of the dramatic developments in this area in recent years.This study collects patent data from the Integrity database. A total of 399 patents with 821 internal citation pairs in the US from 1978 to 2017 were analyzed. Patent citation network analysis was used to visualize the technology relationship.For better production of anti-AD drugs, governments should emphasize the multi-target drug design, provide policy support for private companies, and encourage multilateral cooperation. The β-amyloid peptide (Aβ) theory leaves much to be desired; neurotransmitter and tau protein hypotheses are worth further examination. The use of old drugs for new indications is promising, as are traditional herbal medicines.

Herbal formulae have a long history in clinical medicine in Asia. While the complexity of the formulae leads to the complex compound-target interactions and the resultant multi-target therapeutic effects, it is difficult to elucidate the molecular/therapeutic mechanism of action for the many formulae. For example, the Hua-Yu-Qiang-Shen-Tong-Bi-Fang (TBF), an herbal formula of Chinese medicine, has been used for treating rheumatoid arthritis. However, the target information of a great number of compounds from the TBF formula is missing. In this study, we predicted the targets of the compounds from the TBF formula via network analysis and in silico computing. Initially, the information of the phytochemicals contained in the plants of the herbal formula was collected, and subsequently computed to their corresponding fingerprints for the sake of structural similarity calculation. Then a compound structural similarity network infused with available target information was constructed. Five local similarity indices were used and compared for their performance on predicting the potential new targets of the compounds. Finally, the Preferential Attachment Index was selected for it having an area under curve (AUC) of 0.886, which outperforms the other four algorithms in predicting the compound-target interactions. This method could provide a promising direction for identifying the compound-target interactions of herbal formulae in silico.

S1 An English Version of the Questionnaire. The survey was conducted in Chinese and the original questionnaire was written in Chinese as well. The English version of sample questionnaire may be slightly different to the original Chinese one due to in-house data processing and translation. (DOCX 32 kb)

The U.S. Food and Drug Administration (FDA) approves new drugs every year. Drug targets are some of the most important interactive molecules for drugs, as they have a significant impact on the therapeutic effects of drugs. In this work, we thoroughly analyzed the data of small molecule drugs approved by the U.S. FDA between 2000 and 2015. Specifically, we focused on seven classes of new molecular entity (NME) classified by the anatomic therapeutic chemical (ATC) classification system. They were NMEs and their corresponding targets for the cardiovascular system, respiratory system, nerve system, general anti-infective systemic, genito-urinary system and sex hormones, alimentary tract and metabolisms, and antineoplastic and immunomodulating agents. To study the drug–target interaction on the systems level, we employed network topological analysis and multipartite network projections. As a result, the drug–target relations of different kinds of drugs were comprehensively characterized and global pictures of drug–target, drug–drug, and target–target interactions were visualized and analyzed from the perspective of network models.

Despite the existence of available therapies, the Hepatitis B virus infection continues to be one of the most serious threats to human health, especially in developing countries such as China and India. To shed light on the improvement of current therapies and development of novel anti-HBV drugs, we thoroughly investigated 212 US patents of anti-HBV drugs and analyzed the technology flow in research and development of anti-HBV drugs based on data from IMS LifeCycle databases. Moreover, utilizing the patent citation method, which is an effective indicator of technology flow, we constructed patent citation network models and performed network analysis in order to reveal the features of different technology clusters. As a result, we identified the stagnant status of anti-HBV drug development and pointed the way for development of domestic pharmaceuticals in developing countries. We also discussed about therapeutic vaccines as the potential next generation therapy for HBV infection. Lastly, we depicted the cooperation between entities and found that novel forms of cooperation added diversity to the conventional form of cooperation within the pharmaceutical industry. In summary, our study provides inspiring insights for investors, policy makers, researchers, and other readers interested in anti-HBV drug development.