Efficient Training for Positive Unlabeled Learning

Positive-unlabeled learning identifies vaccine candidate antigens in the malaria parasite Plasmodium falciparum

Malaria vaccine development is hampered by extensive antigenic variation and complex life stages of Plasmodium species. Vaccine development has focused on a small number of antigens, many of which were identified without utilizing systematic genome-level approaches. In this study, we implement a machine learning-based reverse vaccinology approach to predict potential new malaria vaccine candidate antigens. We assemble and analyze P. falciparum proteomic, structural, functional, immunological, genomic, and transcriptomic data, and use positive-unlabeled learning to predict potential antigens based on the properties of known antigens and remaining proteins. We prioritize candidate antigens based on model performance on reference antigens with different genetic diversity and quantify the protein properties that contribute most to identifying top candidates. Candidate antigens are characterized by gene essentiality, gene ontology, and gene expression in different life stages to inform future vaccine development. This approach provides a framework for identifying and prioritizing candidate vaccine antigens for a broad range of pathogens.

Building Fake Review Detection Model Based on Sentiment Intensity and PU Learning

Fake review detection has the characteristics of huge stream data processing scale, unlimited data increment, dynamic change, and so on. However, the existing fake review detection methods mainly target limited and static review data. In addition, deceptive fake reviews have always been a difficult point in fake review detection due to their hidden and diverse characteristics. To solve the above problems, this article proposes a fake review detection model based on sentiment intensity and PU learning (SIPUL), which can continuously learn the prediction model from the constantly arriving streaming data. First, when the streaming data arrive, the sentiment intensity is introduced to divide the reviews into different subsets (i.e., strong sentiment set and weak sentiment set). Then, the initial positive and negative samples are extracted from the subset using the marking mechanism of selection completely at random (SCAR) and Spy technology. Second, building a semi-supervised positive-unlabeled (PU) learning detector based on the initial sample to detect fake reviews in the data stream iteratively. According to the detection results, the data of initial samples and the PU learning detector are continuously updated. Finally, the old data are continually deleted according to the historical record points, so that the training sample data are within a manageable size and prevent overfitting. Experimental results show that the model can effectively detect fake reviews, especially deceptive reviews.

Inferring drug-disease associations by a deep analysis on drug and disease networks

Drugs, which treat various diseases, are essential for human health. However, developing new drugs is quite laborious, time-consuming, and expensive. Although investments into drug development have greatly increased over the years, the number of drug approvals each year remain quite low. Drug repositioning is deemed an effective means to accelerate the procedures of drug development because it can discover novel effects of existing drugs. Numerous computational methods have been proposed in drug repositioning, some of which were designed as binary classifiers that can predict drug-disease associations (DDAs). The negative sample selection was a common defect of this method. In this study, a novel reliable negative sample selection scheme, named RNSS, is presented, which can screen out reliable pairs of drugs and diseases with low probabilities of being actual DDAs. This scheme considered information from k-neighbors of one drug in a drug network, including their associations to diseases and the drug. Then, a scoring system was set up to evaluate pairs of drugs and diseases. To test the utility of the RNSS, three classic classification algorithms (random forest, bayes network and nearest neighbor algorithm) were employed to build classifiers using negative samples selected by the RNSS. The cross-validation results suggested that such classifiers provided a nearly perfect performance and were significantly superior to those using some traditional and previous negative sample selection schemes.

Accurate prediction of virus-host protein-protein interactions via a Siamese neural network using deep protein sequence embeddings

Prediction and understanding of virus-host protein-protein interactions (PPIs) have relevance for the development of novel therapeutic interventions. In addition, virus-like particles open novel opportunities to deliver therapeutics to targeted cell types and tissues. Given our incomplete knowledge of PPIs on the one hand and the cost and time associated with experimental procedures on the other, we here propose a deep learning approach to predict virus-host PPIs. Our method (Siamese Tailored deep sequence Embedding of Proteins [STEP]) is based on recent deep protein sequence embedding techniques, which we integrate into a Siamese neural network. After showing the state-of-the-art performance of STEP on external datasets, we apply it to two use cases, severe acute respiratory syndrome coronavirus 2 and John Cunningham polyomavirus, to predict virus-host PPIs. Altogether our work highlights the potential of deep sequence embedding techniques originating from the field of NLP as well as explainable artificial intelligence methods for the analysis of biological sequences.

Deep Anomaly Detection Based on Variational Deviation Network

There is relatively little research on deep learning for anomaly detection within the field of deep learning. Existing deep anomaly detection methods focus on the learning of feature reconstruction, but such methods mainly learn new feature representations, and the new features do not fully reflect the original features, leading to inaccurate anomaly scores; in addition, there is an end-to-end deep anomaly detection algorithm, but the method cannot accurately obtain a reference score that matches the data themselves. In addition, in most practical scenarios, the data are unlabeled, and there exist some datasets with labels, but the confidence and accuracy of the labels are very low, resulting in inaccurate results when put into the model, which makes them often designed for unsupervised learning, and thus in such algorithms, the prior knowledge of known anomalous data is often not used to optimize the anomaly scores. To address the two problems raised above, this paper proposes a new anomaly detection model that learns anomaly scores mainly through a variational deviation network (i.e., not by learning the reconstruction error of new features, distance metrics, or random generation, but by learning the normal distribution of normal data). In this model, we force the anomaly scores to deviate significantly from the normal data by a small amount of anomalous data and a reference score generated by variational self-encoding. The experimental results in multiple classes of data show that the new variational deviation network proposed in this paper has higher accuracy among the mainstream anomaly detection algorithms.

springD2A: capturing uncertainty in disease-drug association prediction with model integration

MOTIVATION

Drug repositioning that aims to find new indications for existing drugs has been an efficient strategy for drug discovery. In the scenario where we only have confirmed disease-drug associations as positive pairs, a negative set of disease-drug pairs is usually constructed from the unknown disease-drug pairs in previous studies, where we do not know whether drugs and diseases can be associated, to train a model for disease-drug association prediction (drug repositioning). Drugs and diseases in these negative pairs can potentially be associated, but most studies have ignored them.

RESULTS

We present a method, springD2A, to capture the uncertainty in the negative pairs, and to discriminate between positive and unknown pairs because the former are more reliable. In springD2A, we introduce a spring-like penalty for the loss of negative pairs, which is strong if they are too close in a unit sphere, but mild if they are at a moderate distance. We also design a sequential sampling in which the probability of an unknown disease-drug pair sampled as negative is proportional to its score predicted as positive. Multiple models are learned during sequential sampling, and we adopt parameter- and feature-based ensemble schemes to boost performance. Experiments show springD2A is an effective tool for drug-repositioning.

AVAILABILITY AND IMPLEMENTATION

A python implementation of springD2A and datasets used in this study are available at https://github.com/wangyuanhao/springD2A.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Embryo Grading With Unreliable Labels Due to Chromosome Abnormalities by Regularized PU Learning With Ranking

We propose a method for human embryo grading with its images. This grading has been achieved by positive-negative classification (i.e., live birth or non-live birth). However, negative (non-live birth) labels collected in clinical practice are unreliable because the visual features of negative images are equal to those of positive (live birth) images if these non-live birth embryos have chromosome abnormalities. For alleviating an adverse effect of these unreliable labels, our method employs Positive-Unlabeled (PU) learning so that live birth and non-live birth are labeled as positive and unlabeled, respectively, where unlabeled samples contain both positive and negative samples. In our method, this PU learning on a deep CNN is improved by a learning-to-rank scheme. While the original learning-to-rank scheme is designed for positive-negative learning, it is extended to PU learning. Furthermore, overfitting in this PU learning is alleviated by regularization with mutual information. Experimental results with 643 time-lapse image sequences demonstrate the effectiveness of our framework in terms of the recognition accuracy and the interpretability. In quantitative comparison, the full version of our proposed method outperforms positive-negative classification in recall and F-measure by a wide margin (0.22 vs. 0.69 in recall and 0.27 vs. 0.42 in F-measure). In qualitative evaluation, visual attentions estimated by our method are interpretable in comparison with morphological assessments in clinical practice.

GuiltyTargets: Prioritization of Novel Therapeutic Targets With Network Representation Learning

The majority of clinical trials fail due to low efficacy of investigated drugs, often resulting from a poor choice of target protein. Existing computational approaches aim to support target selection either via genetic evidence or by putting potential targets into the context of a disease specific network reconstruction. The purpose of this work was to investigate whether network representation learning techniques could be used to allow for a machine learning based prioritization of putative targets. We propose a novel target prioritization approach, GuiltyTargets, which relies on attributed network representation learning of a genome-wide protein-protein interaction network annotated with disease-specific differential gene expression and uses positive-unlabeled (PU) machine learning for candidate ranking. We evaluated our approach on 12 datasets from six diseases of different type (cancer, metabolic, neurodegenerative) within a 10 times repeated 5-fold stratified cross-validation and achieved AUROC values between 0.92 - 0.97, significantly outperforming previous approaches that relied on manually engineered topological features. Moreover, we showed that GuiltyTargets allows for target repositioning across related disease areas. An application of GuiltyTargets to Alzheimer's disease resulted in a number of highly ranked candidates that are currently discussed as targets in the literature. Interestingly, one (COMT) is also the target of an approved drug (Tolcapone) for Parkinson's disease, highlighting the potential for target repositioning with our method. The GuiltyTargets Python package is available on PyPI and all code used for analysis can be found under the MIT License at https://github.com/GuiltyTargets. Attributed network representation learning techniques provide an interesting approach to effectively leverage the existing knowledge about the molecular mechanisms in different diseases. In this work, the combination with positive-unlabeled learning for target prioritization demonstrated a clear superiority compared to classical feature engineering approaches. Our work highlights the potential of attributed network representation learning for target prioritization. Given the overarching relevance of networks in computational biology we believe that attributed network representation learning techniques could have a broader impact in the future.

Instance-Dependent Positive and Unlabeled Learning with Labeling Bias Estimation

This paper studies instance-dependent Positive and Unlabeled (PU) classification, where whether a positive example will be labeled (indicated by s) is not only related to the class label y, but also depends on the observation x. Therefore, the labeling probability on positive examples is not uniform as previous works assumed, but is biased to some simple or critical data points. To depict the above dependency relationship, a graphical model is built in this paper which further leads to a maximization problem on the induced likelihood function regarding P(s,y|x). By utilizing the well-known EM and Adam optimization techniques, the labeling probability of any positive example P(s=1|y=1,x) as well as the classifier induced by P(y|x) can be acquired. Theoretically, we prove that the critical solution always exists, and is locally unique for linear model if some sufficient conditions are met. Moreover, we upper bound the generalization error for both linear logistic and non-linear network instantiations of our algorithm. Empirically, we compare our method with state-of-the-art instance-independent and instance-dependent PU algorithms on a wide range of synthetic, benchmark and real-world datasets, and the experimental results firmly demonstrate the advantage of the proposed method over the existing PU approaches.

Probing lncRNA-Protein Interactions: Data Repositories, Models, and Algorithms

Identifying lncRNA-protein interactions (LPIs) is vital to understanding various key biological processes. Wet experiments found a few LPIs, but experimental methods are costly and time-consuming. Therefore, computational methods are increasingly exploited to capture LPI candidates. We introduced relevant data repositories, focused on two types of LPI prediction models: network-based methods and machine learning-based methods. Machine learning-based methods contain matrix factorization-based techniques and ensemble learning-based techniques. To detect the performance of computational methods, we compared parts of LPI prediction models on Leave-One-Out cross-validation (LOOCV) and fivefold cross-validation. The results show that SFPEL-LPI obtained the best performance of AUC. Although computational models have efficiently unraveled some LPI candidates, there are many limitations involved. We discussed future directions to further boost LPI predictive performance.

Revealing Drug-Target Interactions with Computational Models and Algorithms

Background: Identifying possible drug-target interactions (DTIs) has become an important task in drug research and development. Although high-throughput screening is becoming available, experimental methods narrow down the validation space because of extremely high cost, low success rate, and time consumption. Therefore, various computational models have been exploited to infer DTI candidates. Methods: We introduced relevant databases and packages, mainly provided a comprehensive review of computational models for DTI identification, including network-based algorithms and machine learning-based methods. Specially, machine learning-based methods mainly include bipartite local model, matrix factorization, regularized least squares, and deep learning. Results: Although computational methods have obtained significant improvement in the process of DTI prediction, these models have their limitations. We discussed potential avenues for boosting DTI prediction accuracy as well as further directions.