Global discriminative learning for higher-accuracy computational gene prediction

Most ab initio gene predictors use a probabilistic sequence model, typically a hidden Markov model, to combine separately trained models of genomic signals and content. By combining separate models of relevant genomic features, such gene predictors can exploit small training sets and incomplete annotations, and can be trained fairly efficiently. However, that type of piecewise training does not optimize prediction accuracy and has difficulty in accounting for statistical dependencies among different parts of the gene model. With genomic information being created at an ever-increasing rate, it is worth investigating alternative approaches in which many different types of genomic evidence, with complex statistical dependencies, can be integrated by discriminative learning to maximize annotation accuracy. Among discriminative learning methods, large-margin classifiers have become prominent because of the success of support vector machines (SVM) in many classification tasks. We describe CRAIG, a new program for ab initio gene prediction based on a conditional random field model with semi-Markov structure that is trained with an online large-margin algorithm related to multiclass SVMs. Our experiments on benchmark vertebrate datasets and on regions from the ENCODE project show significant improvements in prediction accuracy over published gene predictors that use intrinsic features only, particularly at the gene level and on genes with long introns.     

Small, fuzzy and interpretable gene expression based classifiers

MOTIVATION: Interpretation of classification models derived from gene-expression data is usually not simple, yet it is an important aspect in the analytical process. We investigate the performance of small rule-based classifiers based on fuzzy logic in five datasets that are different in size, laboratory origin and biomedical domain. RESULTS: The classifiers resulted in rules that can be readily examined by biomedical researchers. The fuzzy-logic-based classifiers compare favorably with logistic regression in all datasets. AVAILABILITY: Prototype available upon request.     

Biomarker discovery across annotated and unannotated microarray datasets using semi-supervised learning

The growing body of DNA microarray data has the potential to advance our understanding of the molecular basis of disease. However annotating microarray datasets with clinically useful information is not always possible, as this often requires access to detailed patient records. In this study we introduce GLAD, a new Semi-Supervised Learning (SSL) method for combining independent annotated datasets and unannotated datasets with the aim of identifying more robust sample classifiers. In our method, independent models are developed using subsets of genes for the annotated and unannotated datasets. These models are evaluated according to a scoring function that incorporates terms for classification accuracy on annotated data, and relative cluster separation in unannotated data. Improved models are iteratively generated using a genetic algorithm feature selection technique. Our results show that the addition of unannotated data into training, significantly improves classifier robustness.     

A hidden Markov model-based stride segmentation technique applied to equine inertial sensor trunk movement data

Inertial sensors are now sufficiently small and lightweight to be used for the collection of large datasets of both humans and animals. However, processing of these large datasets requires a certain degree of automation to achieve realistic workloads. Hidden Markov models (HMMs) are widely used stochastic pattern recognition tools and enable classification of non-stationary data. Here we apply HMMs to identify and segment into strides, data collected from a trunk-mounted six degrees of freedom inertial sensor in galloping Thoroughbred racehorses. A data set comprising mixed gait sequences from seven horses was subdivided into training, cross-validation and independent test set. Manual gallop stride segmentations were created and used for training as well as for evaluating cross-validation and test set performance. On the test set, 91% of the strides were accurately detected to lie within +/- 40 ms (< 10% stride time) of the manually segmented stride starts. While the automated system did not miss any of the strides, it identified additional gallop strides at the beginning of the trials. In the light of increasing use of inertial sensors for ambulatory measurements in clinical settings, automated processing techniques will be required for efficient data processing to enable instantaneous decision making from large amounts of data. In this context, automation is essential to gain optimal benefits from the potentially increased statistical power associated with large numbers of strides that can be collected in a relatively short period of time. We propose the use of HMM-based classifiers since they are easy to implement. In the present study, consistent results across cross-validation and test set were achieved with limited training data.     

Tag Frequency Difference: Rapid estimation of image set relevance for species occurrence data using general-purpose image classifiers

iEcology is used to supplement traditional ecological data by sourcing large quantities of media from the internet. Images and their metadata are widely available online and can provide information on species occurrence, behaviour and visible traits. However, this data is inherently noisy and data quality varies significantly between sources. Many iEcology studies utilise data from a single source for simplicity and efficiency. Hence, a tool to compare the suitability of different media sources in addressing a particular research question is needed.We provide a simple, novel way to estimate the fraction of images within multiple unverified datasets that potentially depict a specified target fauna. Our method, the Sum of Tag Frequency Differences (STFD), uses any pretrained, general-purpose image classifier. One of the method's innovations is that it does not require training the classifier to recognise the target fauna. Instead, STFD analyses the frequency of the generic text-tags returned by a classifier for multiple datasets and compares them to the corresponding frequencies of an authoritative image dataset that depicts only the target organism. From this comparison, STFD allows us to deduce the fraction of images of the target in unverified datasets.To validate the STFD approach, we processed images from five sources: Flickr, iNaturalist, Instagram, Reddit and Twitter. For each media source, we conducted an STFD analysis of three fauna invasive to Australia: Cane toads (Rhinella marina), German wasps (Vespula germanica), and the higher-level colloquial taxonomic classification, “wild rabbits”. We found the STFD provided an accurate assessment of image source relevance across all data sources and target organisms. This was demonstrated by the consistent, very strong correlation (toads r ≥0.97, wasps r ≥0.95, wild rabbits≥ 0.95) between STFD predictions, and the fraction of target images in a source dataset observed by a human expert.The STFD provides a low-cost, simple and accurate comparison of the relevance of online image sources to specific fauna for iEcology applications. It does not require expertise in machine learning or training neural-network species-specific classifiers. The method enables researchers to assess multiple image sources to select those warranting detailed investigation for the development of tools for web-scraping, citizen science campaigns, further monitoring or analysis.     

Gene Function Prediction Based on the Gene Ontology Hierarchical Structure

The information of the Gene Ontology annotation is helpful in the explanation of life science phenomena, and can provide great support for the research of the biomedical field. The use of the Gene Ontology is gradually affecting the way people store and understand bioinformatic data. To facilitate the prediction of gene functions with the aid of text mining methods and existing resources, we transform it into a multi-label top-down classification problem and develop a method that uses the hierarchical relationships in the Gene Ontology structure to relieve the quantitative imbalance of positive and negative training samples. Meanwhile the method enhances the discriminating ability of classifiers by retaining and highlighting the key training samples. Additionally, the top-down classifier based on a tree structure takes the relationship of target classes into consideration and thus solves the incompatibility between the classification results and the Gene Ontology structure. Our experiment on the Gene Ontology annotation corpus achieves an F-value performance of 50.7% (precision: 52.7% recall: 48.9%). The experimental results demonstrate that when the size of training set is small, it can be expanded via topological propagation of associated documents between the parent and child nodes in the tree structure. The top-down classification model applies to the set of texts in an ontology structure or with a hierarchical relationship.     

Classification of deadlift biomechanics with wearable inertial measurement units

The deadlift is a compound full-body exercise that is fundamental in resistance training, rehabilitation programs and powerlifting competitions. Accurate quantification of deadlift biomechanics is important to reduce the risk of injury and ensure training and rehabilitation goals are achieved. This study sought to develop and evaluate deadlift exercise technique classification systems utilising Inertial Measurement Units (IMUs), recording at 51.2 Hz, worn on the lumbar spine, both thighs and both shanks. It also sought to compare classification quality when these IMUs are worn in combination and in isolation. Two datasets of IMU deadlift data were collected. Eighty participants first completed deadlifts with acceptable technique and 5 distinct, deliberately induced deviations from acceptable form. Fifty-five members of this group also completed a fatiguing protocol (3-Repition Maximum test) to enable the collection of natural deadlift deviations. For both datasets, universal and personalised random-forests classifiers were developed and evaluated. Personalised classifiers outperformed universal classifiers in accuracy, sensitivity and specificity in the binary classification of acceptable or aberrant technique and in the multi-label classification of specific deadlift deviations. Whilst recent research has favoured universal classifiers due to the reduced overhead in setting them up for new system users, this work demonstrates that such techniques may not be appropriate for classifying deadlift technique due to the poor accuracy achieved. However, personalised classifiers perform very well in assessing deadlift technique, even when using data derived from a single lumbar-worn IMU to detect specific naturally occurring technique mistakes.     

Partition dataset according to amino acid type improves the prediction of deleterious non-synonymous SNPs

Many non-synonymous SNPs (nsSNPs) are associated with diseases, and numerous machine learning methods have been applied to train classifiers for sorting disease-associated nsSNPs from neutral ones. The continuously accumulated nsSNP data allows us to further explore better prediction approaches. In this work, we partitioned the training data into 20 subsets according to either original or substituted amino acid type at the nsSNP site. Using support vector machine (SVM), training classification models on each subset resulted in an overall accuracy of 76.3% or 74.9% depending on the two different partition criteria, while training on the whole dataset obtained an accuracy of only 72.6%. Moreover, the dataset was also randomly divided into 20 subsets, but the corresponding accuracy was only 73.2%. Our results demonstrated that partitioning the whole training dataset into subsets properly, i.e., according to the residue type at the nsSNP site, will improve the performance of the trained classifiers significantly, which should be valuable in developing better tools for predicting the disease-association of nsSNPs.     

Shark detection and classification with machine learning

Suitable shark conservation depends on well-informed population assessments. Direct methods such as scientific surveys and fisheries monitoring are adequate for defining population statuses, but species-specific indices of abundance and distribution coming from these sources are rare for most shark species. We can rapidly fill these information gaps by boosting media-based remote monitoring efforts with machine learning and automation.We created a database of 53,345 shark images covering 219 species of sharks, and packaged object-detection and image classification models into a Shark Detector bundle. The Shark Detector recognizes and classifies sharks from videos and images using transfer learning and convolutional neural networks (CNNs). We applied these models to common data-generation approaches of sharks: collecting occurrence records from photographs taken by the public or citizen scientists, processing baited remote camera footage and online videos, and data-mining Instagram. We examined the accuracy of each model and tested genus and species prediction correctness as a result of training data quantity.The Shark Detector can classify 47 species pertaining to 26 genera. It sorted heterogeneous datasets of images sourced from Instagram with 91% accuracy and classified species with 70% accuracy. It located sharks in baited remote footage and YouTube videos with 89% accuracy, and classified located subjects to the species level with 69% accuracy. All data-generation methods were processed without manual interaction.As media-based remote monitoring appears to dominate methods for observing sharks in nature, we developed an open-source Shark Detector to facilitate common identification applications. Prediction accuracy of the software pipeline increases as more images are added to the training dataset. We provide public access to the software on our GitHub page.     

Prediction of cell penetrating peptides by support vector machines

Cell penetrating peptides (CPPs) are those peptides that can transverse cell membranes to enter cells. Once inside the cell, different CPPs can localize to different cellular components and perform different roles. Some generate pore-forming complexes resulting in the destruction of cells while others localize to various organelles. Use of machine learning methods to predict potential new CPPs will enable more rapid screening for applications such as drug delivery. We have investigated the influence of the composition of training datasets on the ability to classify peptides as cell penetrating using support vector machines (SVMs). We identified 111 known CPPs and 34 known non-penetrating peptides from the literature and commercial vendors and used several approaches to build training data sets for the classifiers. Features were calculated from the datasets using a set of basic biochemical properties combined with features from the literature determined to be relevant in the prediction of CPPs. Our results using different training datasets confirm the importance of a balanced training set with approximately equal number of positive and negative examples. The SVM based classifiers have greater classification accuracy than previously reported methods for the prediction of CPPs, and because they use primary biochemical properties of the peptides as features, these classifiers provide insight into the properties needed for cell-penetration. To confirm our SVM classifications, a subset of peptides classified as either penetrating or non-penetrating was selected for synthesis and experimental validation. Of the synthesized peptides predicted to be CPPs, 100% of these peptides were shown to be penetrating.     

应用生长、分级的SOM模型进行意识任务分类

提出一种使用生长、分级的自组织映射（growing hierarchical self-organizing map,GHSOM）模型进行基于EEG信号的意识任务分类来实现脑机接口技术的方法。GHSOM模型是自组织映射（self-organizing map,SOM）的一种变体,由多层的SOM组成,具有一定的分级结构,能够表达数据中不同层次的信息。同时研究了使用平均量化误差（mean quantization error,mqe）和量化误差（quantization error,qe）两种方法实现的GHSOM模型对意识任务分类的作用。结果表明,GHSOM模型对于意识任务的可分性能够提供可视化的信息,并且发现使用量化误差方法实现的GHSOM模型提供较多的数据信息和较高的分类精度。使用GHSOM模型进行了5类意识任务的分类,平均分类精度可达80％。     

A comparative study of feature selection and multiclass classification methods for tissue classification based on gene expression

This paper studies the problem of building multiclass classifiers for tissue classification based on gene expression. The recent development of microarray technologies has enabled biologists to quantify gene expression of tens of thousands of genes in a single experiment. Biologists have begun collecting gene expression for a large number of samples. One of the urgent issues in the use of microarray data is to develop methods for characterizing samples based on their gene expression. The most basic step in the research direction is binary sample classification, which has been studied extensively over the past few years. This paper investigates the next step-multiclass classification of samples based on gene expression. The characteristics of expression data (e.g. large number of genes with small sample size) makes the classification problem more challenging. The process of building multiclass classifiers is divided into two components: (i) selection of the features (i.e. genes) to be used for training and testing and (ii) selection of the classification method. This paper compares various feature selection methods as well as various state-of-the-art classification methods on various multiclass gene expression datasets. Our study indicates that multiclass classification problem is much more difficult than the binary one for the gene expression datasets. The difficulty lies in the fact that the data are of high dimensionality and that the sample size is small. The classification accuracy appears to degrade very rapidly as the number of classes increases. In particular, the accuracy was very low regardless of the choices of the methods for large-class datasets (e.g. NCI60 and GCM). While increasing the number of samples is a plausible solution to the problem of accuracy degradation, it is important to develop algorithms that are able to analyze effectively multiple-class expression data for these special datasets.     

Predicting Classifier Performance with Limited Training Data: Applications to Computer-Aided Diagnosis in Breast and Prostate Cancer

Clinical trials increasingly employ medical imaging data in conjunction with supervised classifiers, where the latter require large amounts of training data to accurately model the system. Yet, a classifier selected at the start of the trial based on smaller and more accessible datasets may yield inaccurate and unstable classification performance. In this paper, we aim to address two common concerns in classifier selection for clinical trials: (1) predicting expected classifier performance for large datasets based on error rates calculated from smaller datasets and (2) the selection of appropriate classifiers based on expected performance for larger datasets. We present a framework for comparative evaluation of classifiers using only limited amounts of training data by using random repeated sampling (RRS) in conjunction with a cross-validation sampling strategy. Extrapolated error rates are subsequently validated via comparison with leave-one-out cross-validation performed on a larger dataset. The ability to predict error rates as dataset size increases is demonstrated on both synthetic data as well as three different computational imaging tasks: detecting cancerous image regions in prostate histopathology, differentiating high and low grade cancer in breast histopathology, and detecting cancerous metavoxels in prostate magnetic resonance spectroscopy. For each task, the relationships between 3 distinct classifiers (k-nearest neighbor, naive Bayes, Support Vector Machine) are explored. Further quantitative evaluation in terms of interquartile range (IQR) suggests that our approach consistently yields error rates with lower variability (mean IQRs of 0.0070, 0.0127, and 0.0140) than a traditional RRS approach (mean IQRs of 0.0297, 0.0779, and 0.305) that does not employ cross-validation sampling for all three datasets.     

A comparison of supervised learning techniques in the classification of bat echolocation calls

Today's acoustic monitoring devices are capable of recording and storing tremendous amounts of data. Until recently, the classification of animal vocalizations from field recordings has been relegated to qualitative approaches. For large-scale acoustic monitoring studies, qualitative approaches are very time-consuming and suffer from the bias of subjectivity. Recent developments in supervised learning techniques can provide rapid, accurate, species-level classification of bioacoustics data. We compared the classification performances of four supervised learning techniques (random forests, support vector machines, artificial neural networks, and discriminant function analysis) for five different classification tasks using bat echolocation calls recorded by a popular frequency-division bat detector. We found that all classifiers performed similarly in terms of overall accuracy with the exception of discriminant function analysis, which had the lowest average performance metrics. Random forests had the advantage of high sensitivities, specificities, and predictive powers across the majority of classification tasks, and also provided metrics for determining the relative importance of call features in distinguishing between groups. Overall classification accuracy for each task was slightly lower than reported accuracies using calls recorded by time-expansion detectors. Myotis spp. were particularly difficult to separate; classifiers performed best when members of this genus were combined in genus-level classification and analyzed separately at the level of species. Additionally, we identified and ranked the relative contributions of all predictor features to classifier accuracy and found measurements of frequency, total call duration, and characteristic slope to be the most important contributors to classification success. We provide recommendations to maximize accuracy and efficiency when analyzing acoustic data, and suggest an application of automated bioacoustics monitoring to contribute to wildlife monitoring efforts.     

Multiclass cancer classification by support vector machines with class-wise optimized genes and probability estimates

We investigate the multiclass classification of cancer microarray samples. In contrast to classification of two cancer types from gene expression data, multiclass classification of more than two cancer types are relatively hard and less studied problem. We used class-wise optimized genes with corresponding one-versus-all support vector machine (OVA-SVM) classifier to maximize the utilization of selected genes. Final prediction was made by using probability scores from all classifiers. We used three different methods of estimating probability from decision value. Among the three probability methods, Platt's approach was more consistent, whereas, isotonic approach performed better for datasets with unequal proportion of samples in different classes. Probability based decision does not only gives true and fair comparison between different one-versus-all (OVA) classifiers but also gives the possibility of using them for any post analysis. Several ensemble experiments, an example of post analysis, of the three probability methods were implemented to study their effect in improving the classification accuracy. We observe that ensemble did help in improving the predictive accuracy of cancer data sets especially involving unbalanced samples. Four-fold external stratified cross-validation experiment was performed on the six multiclass cancer datasets to obtain unbiased estimates of prediction accuracies. Analysis of class-wise frequently selected genes on two cancer datasets demonstrated that the approach was able to select important and relevant genes consistent to literature. This study demonstrates successful implementation of the framework of class-wise feature selection and multiclass classification for prediction of cancer subtypes on six datasets.     

Extraction of Pharmacokinetic Evidence of Drug–Drug Interactions from the Literature

Drug-drug interaction (DDI) is a major cause of morbidity and mortality and a subject of intense scientific interest. Biomedical literature mining can aid DDI research by extracting evidence for large numbers of potential interactions from published literature and clinical databases. Though DDI is investigated in domains ranging in scale from intracellular biochemistry to human populations, literature mining has not been used to extract specific types of experimental evidence, which are reported differently for distinct experimental goals. We focus on pharmacokinetic evidence for DDI, essential for identifying causal mechanisms of putative interactions and as input for further pharmacological and pharmacoepidemiology investigations. We used manually curated corpora of PubMed abstracts and annotated sentences to evaluate the efficacy of literature mining on two tasks: first, identifying PubMed abstracts containing pharmacokinetic evidence of DDIs; second, extracting sentences containing such evidence from abstracts. We implemented a text mining pipeline and evaluated it using several linear classifiers and a variety of feature transforms. The most important textual features in the abstract and sentence classification tasks were analyzed. We also investigated the performance benefits of using features derived from PubMed metadata fields, various publicly available named entity recognizers, and pharmacokinetic dictionaries. Several classifiers performed very well in distinguishing relevant and irrelevant abstracts (reaching F1≈0.93, MCC≈0.74, iAUC≈0.99) and sentences (F1≈0.76, MCC≈0.65, iAUC≈0.83). We found that word bigram features were important for achieving optimal classifier performance and that features derived from Medical Subject Headings (MeSH) terms significantly improved abstract classification. We also found that some drug-related named entity recognition tools and dictionaries led to slight but significant improvements, especially in classification of evidence sentences. Based on our thorough analysis of classifiers and feature transforms and the high classification performance achieved, we demonstrate that literature mining can aid DDI discovery by supporting automatic extraction of specific types of experimental evidence.     

R package for animal behavior classification from accelerometer data—rabc

Increasingly, animal behavior studies are enhanced through the use of accelerometry. To allow translation of raw accelerometer data to animal behaviors requires the development of classifiers. Here, we present the “rabc” (r for animal behavior classification) package to assist researchers with the interactive development of such animal behavior classifiers in a supervised classification approach. The package uses datasets consisting of accelerometer data with their corresponding animal behaviors (e.g., for triaxial accelerometer data along the x, y and z axes arranged as “x, y, z, x, y, z,…, behavior”). Using an example dataset collected on white stork (Ciconia ciconia), we illustrate the workflow of this package, including accelerometer data visualization, feature calculation, feature selection, feature visualization, extreme gradient boost model training, validation, and, finally, a demonstration of the behavior classification results.