Twan van Laarhoven

Gaussian interaction profile kernels for predicting drug–target interaction

MOTIVATION The in silico prediction of potential interactions between drugs and target proteins is of core importance for the identification of new drugs or novel targets for existing drugs. However, only a tiny portion of all drug-target pairs in current datasets are experimentally validated interactions. This motivates the need for developing computational methods that predict true interaction pairs with high accuracy. RESULTS We show that a simple machine learning method that uses the drug-target network as the only source of information is capable of predicting true interaction pairs with high accuracy. Specifically, we introduce interaction profiles of drugs (and of targets) in a network, which are binary vectors specifying the presence or absence of interaction with every target (drug) in that network. We define a kernel on these profiles, called the Gaussian Interaction Profile (GIP) kernel, and use a simple classifier, (kernel) Regularized Least Squares (RLS), for prediction drug-target interactions. We test comparatively the effectiveness of RLS with the GIP kernel on four drug-target interaction networks used in previous studies. The proposed algorithm achieves area under the precision-recall curve (AUPR) up to 92.7, significantly improving over results of state-of-the-art methods. Moreover, we show that using also kernels based on chemical and genomic information further increases accuracy, with a neat improvement on small datasets. These results substantiate the relevance of the network topology (in the form of interaction profiles) as source of information for predicting drug-target interactions. AVAILABILITY Software and Supplementary Material are available at http://cs.ru.nl/~tvanlaarhoven/drugtarget2011/. CONTACT tvanlaarhoven@cs.ru.nl; elenam@cs.ru.nl. SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Overview and evaluation of premise selection techniques for large theory mathematics

In this paper, an overview of state-of-the-art techniques for premise selection in large theory mathematics is provided, and new premise selection techniques are introduced. Several evaluation metrics are introduced, compared and their appropriateness is discussed in the context of automated reasoning in large theory mathematics. The methods are evaluated on the MPTP2078 benchmark, a subset of the Mizar library, and a 10% improvement is obtained over the best method so far.

Robust community detection methods with resolution parameter for complex detection in protein protein interaction networks

Unraveling the community structure of real-world networks is an important and challenging problem. Recently, it has been shown that methods based on optimizing a clustering measure, in particular modularity, have a resolution bias, e.g. communities with sizes below some threshold remain unresolved. This problem has been tackled by incorporating a parameter in the method which influences the size of the communities. Methods incorporating this type of parameter are also called multi-resolution methods. In this paper we consider fast greedy local search optimization of a clustering objective function with two different objective functions incorporating a resolution parameter: modularity and a function we introduced in a recent work, called w-log-v. We analyze experimentally the performance of the resulting algorithms when applied to protein-protein interaction (PPI) networks. Specifically, publicly available yeast protein networks from past studies, as well as the present BioGRID database, are considered. Furthermore, to test robustness of the methods, various types of randomly perturbed networks obtained from the BioGRID data are also considered. Results of extensive experiments show improved or competitive performance over MCL, a state-of-the-art algorithm for complex detection in PPI networks, in particular on BioGRID data, where w-log-v obtains excellent accuracy and robustness performance.

Deep learning for automatic stereotypical motor movement detection using wearable sensors in autism spectrum disorders

Autism Spectrum Disorders are associated with atypical movements, of which stereotypical motor movements (SMMs) interfere with learning and social interaction. The automatic SMM detection using inertial measurement units (IMU) remains complex due to the strong intra and inter-subject variability, especially when handcrafted features are extracted from the signal. We propose a new application of the deep learning to facilitate automatic SMM detection using multi-axis IMUs. We use a convolutional neural network (CNN) to learn a discriminative feature space from raw data. We show how the CNN can be used for parameter transfer learning to enhance the detection rate on longitudinal data. We also combine the long short-term memory (LSTM) with CNN to model the temporal patterns in a sequence of multi-axis signals. Further, we employ ensemble learning to combine multiple LSTM learners into a more robust SMM detector. Our results show that: 1) feature learning outperforms handcrafted features; 2) parameter transfer learning is beneficial in longitudinal settings; 3) using LSTM to learn the temporal dynamic of signals enhances the detection rate especially for skewed training data; 4) an ensemble of LSTMs provides more accurate and stable detectors. These findings provide a significant step toward accurate SMM detection in real-time scenarios.

Novelty Detection using Deep Normative Modeling for IMU-Based Abnormal Movement Monitoring in Parkinson’s Disease and Autism Spectrum Disorders

Detecting and monitoring of abnormal movement behaviors in patients with Parkinson’s Disease (PD) and individuals with Autism Spectrum Disorders (ASD) are beneficial for adjusting care and medical treatment in order to improve the patient’s quality of life. Supervised methods commonly used in the literature need annotation of data, which is a time-consuming and costly process. In this paper, we propose deep normative modeling as a probabilistic novelty detection method, in which we model the distribution of normal human movements recorded by wearable sensors and try to detect abnormal movements in patients with PD and ASD in a novelty detection framework. In the proposed deep normative model, a movement disorder behavior is treated as an extreme of the normal range or, equivalently, as a deviation from the normal movements. Our experiments on three benchmark datasets indicate the effectiveness of the proposed method, which outperforms one-class SVM and the reconstruction-based novelty detection approaches. Our contribution opens the door toward modeling normal human movements during daily activities using wearable sensors and eventually real-time abnormal movement detection in neuro-developmental and neuro-degenerative disorders.

Spectral-Spatial Classification of Hyperspectral Images: Three Tricks and a New Learning Setting

Spectral-spatial classification of remotely sensed hyperspectral images has been the subject of many studies in recent years. Current methods achieve excellent performance on benchmark hyperspectral image labeling tasks when a sufficient number of labeled pixels is available. However, in the presence of only very few labeled pixels, such classification becomes a challenging problem. In this paper we propose to tackle this problem using convolutional neural networks (CNNs) and data augmentation. Our newly developed method relies on the assumption of spectral-spatial locality: nearby pixels in a hyperspectral image are related, in the sense that their spectra and their labels are likely to be similar. We exploit this assumption to develop 1) a new data augmentation procedure which adds new samples to the train set and 2) a tailored loss function which penalize differences among weights of the network corresponding to nearby wavelengths of the spectra. We train a simple single layer convolutional neural network with this loss function and augmented train set and use it to classify all unlabeled pixels of the given image. To assess the efficacy of our method, we used five publicly available hyperspectral images: Pavia Center, Pavia University, KSC, Indian Pines and Salina. On these images our method significantly outperforms other baselines. Notably, with just 1% of labeled pixels per class, on these dataset our method achieves an accuracy of 99.5%, etc. Furthermore we show that our method improves over other baselines also in a supervised setting, when no overlap between train and test pixels is allowed. Overall our investigation demonstrates that spectral-spatial locality can be easily embedded in a simple convolutional neural network through data augmentation and a tailored loss function.Spectral-spatial classification of hyperspectral images has been the subject of many studies in recent years. In the presence of only very few labeled pixels, this task becomes challenging. In this paper we address the following two research questions: 1) Can a simple neural network with just a single hidden layer achieve state of the art performance in the presence of few labeled pixels? 2) How is the performance of hyperspectral image classification methods affected when using disjoint train and test sets? We give a positive answer to the first question by using three tricks within a very basic shallow Convolutional Neural Network (CNN) architecture: a tailored loss function, and smooth- and label-based data augmentation. The tailored loss function enforces that neighborhood wavelengths have similar contributions to the features generated during training. A new label-based technique here proposed favors selection of pixels in smaller classes, which is beneficial in the presence of very few labeled pixels and skewed class distributions. To address the second question, we introduce a new sampling procedure to generate disjoint train and test set. Then the train set is used to obtain the CNN model, which is then applied to pixels in the test set to estimate their labels. We assess the efficacy of the simple neural network method on five publicly available hyperspectral images. On these images our method significantly outperforms considered baselines. Notably, with just 1% of labeled pixels per class, on these datasets our method achieves an accuracy that goes from 86.42% (challenging dataset) to 99.52% (easy dataset). Furthermore we show that the simple neural network method improves over other baselines in the new challenging supervised setting. Our analysis substantiates the highly beneficial effect of using the entire image (so train and test data) for constructing a model.

Generalized Convolution Spectral Mixture for Multi-Task Gaussian Processes

Multi-task Gaussian processes (MTGPs) are a powerful approach for modeling structured dependencies among multiple tasks. Researchers on MTGPs have contributed to enhance this approach in various ways. Current MTGP methods, however, cannot model nonlinear task correlations in a general way. In this paper we address this problem. We focus on spectral mixture (SM) based kernels and propose an enhancement of this type of kernels, called multi-task generalized convolution spectral mixture (MT-GCSM) kernel. The MT-GCSM kernel can model nonlinear task correlations and mixtures dependency, including time and phase delay, not only between different tasks but also within a task at the spectral mixture level. Each task in MT-GCSM has its own generalized convolution spectral mixture kernel (GCSM) with a different number of convolution structures and all spectral mixtures from different tasks are dependent. Furthermore, the proposed kernel uses inner and outer full cross convolution between base spectral mixtures, so that the base spectral mixtures in the tasks are not necessarily aligned. Extensive experiments on synthetic and real-life datasets illustrate the difference between MT-GCSM and other kernels as well as the practical effectiveness of MT-GCSM. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 22 October 2018 doi:10.20944/preprints201810.0461.v1

Pattern Recognition in Bioinformatics

Proteins are known to interact with each other by forming protein complexes and in order to perform specific biological functions. Many community detection methods have been devised for the discovery of protein complexes in protein interaction networks. One common problem in current agglomerative community detection approaches is that vertices with just one neighbor are often classified as separate clusters, which does not make sense for complex identification. Also, a major limitation of agglomerative techniques is that their computational efficiency do not scale well to large protein interaction networks (PINs). In this paper, we propose a new agglomerative algorithm, FAC-PIN, based on a local premetric of relative vertex-to-vertex clustering value and which addresses the above two issues. Our proposed FAC-PIN method is applied to eight PINs from different species, and the identified complexes are validated using experimentally verified complexes. The preliminary computational results show that FAC-PIN can discover protein complexes from PINs more accurately and faster than the HC-PIN and CNM algorithms, the current state-of-the-art agglomerative approaches to complex prediction.