Comparison of subject-independent and subject-specific EEG-based BCI using LDA and SVM classifiers.

Motor imagery brain-computer interface (MI-BCI) is one of the most used paradigms in EEG-based brain-computer interface (BCI). The current state-of-the-art in BCI involves tuning classifiers to subject-specific training data, acquired over several sessions, in order to perform calibration prior to actual use of the so-called subject-specific BCI system (SS-BCI). Herein, the goal is to provide a ready-to-use system requiring minimal effort for setup. Thus, our challenge was to design a subject-independent BCI (SI-BCI) to be used by any new user without the constraint of individual calibration. Outcomes from other studies with the same purpose were used to undertake comparisons and validate our findings. For the EEG signal processing, we used a combination of the delta (0.5-4 Hz), alpha (8-13 Hz), and beta+gamma (13-40 Hz) bands at a stage prior to feature extraction. Next, we extracted features from the 27-channel EEG using common spatial pattern (CSP) and performed binary classification (MI of right- and left-hand) with linear discriminant analysis (LDA) and support vector machine (SVM) classifiers. These analyses were done for both the SS-BCI and SI-BCI models. We employed "leave-one-subject-out" (LOSO) arrangement and 10-fold cross-validation to evaluate our SI-BCI and SS-BCI systems, respectively. Compared with other two studies, our work was the only one that showed higher accuracy for the LDA classifier in SI-BCI as compared to SS-BCI. On the other hand, LDA accuracy was lower than accuracy achieved with SVM in both conditions (SI-BCI and SS-BCI). Our SS-BCI accuracy reached 76.85% using LDA and 94.20% using SVM and for SI-BCI we got 80.30% with LDA and 83.23% with SVM. We conclude that SI-BCI may be a feasible and relevant option, which can be used in scenarios where subjects are not able to submit themselves to long training sessions or to fast evaluation of the so called "BCI illiteracy." Comparatively, our strategy proved to be more efficient, giving us the best result for SI-BCI when faced against the classification performances of other three studies, even considering the caveat that different datasets were used in the comparison of the four studies.

Comparison of Chemometric Problems in Food Analysis Using Non-Linear Methods.

Food analysis is a challenging analytical problem, often addressed using sophisticated laboratory methods that produce large data sets. Linear and non-linear multivariate methods can be used to process these types of datasets and to answer questions such as whether product origin is accurately labeled or whether a product is safe to eat. In this review, we present the application of non-linear methods such as artificial neural networks, support vector machines, self-organizing maps, and multi-layer artificial neural networks in the field of chemometrics related to food analysis. We discuss criteria to determine when non-linear methods are better suited for use instead of traditional methods. The principles of algorithms are described, and examples are presented for solving the problems of exploratory analysis, classification, and prediction.

A Support Vector Machine based method to distinguish long non-coding RNAs from protein coding transcripts.

BACKGROUND: In recent years, a rapidly increasing number of RNA transcripts has been generated by thousands of sequencing projects around the world, creating enormous volumes of transcript data to be analyzed. An important problem to be addressed when analyzing this data is distinguishing between long non-coding RNAs (lncRNAs) and protein coding transcripts (PCTs). Thus, we present a Support Vector Machine (SVM) based method to distinguish lncRNAs from PCTs, using features based on frequencies of nucleotide patterns and ORF lengths, in transcripts. METHODS: The proposed method is based on SVM and uses the first ORF relative length and frequencies of nucleotide patterns selected by PCA as features. FASTA files were used as input to calculate all possible features. These features were divided in two sets: (i) 336 frequencies of nucleotide patterns; and (ii) 4 features derived from ORFs. PCA were applied to the first set to identify 6 groups of frequencies that could most contribute to the distinction. Twenty-four experiments using the 6 groups from the first set and the features from the second set where built to create the best model to distinguish lncRNAs from PCTs. RESULTS: This method was trained and tested with human (Homo sapiens), mouse (Mus musculus) and zebrafish (Danio rerio) data, achieving 98.21%, 98.03% and 96.09%, accuracy, respectively. Our method was compared to other tools available in the literature (CPAT, CPC, iSeeRNA, lncRNApred, lncRScan-SVM and FEELnc), and showed an improvement in accuracy by ≈3.00%. In addition, to validate our model, the mouse data was classified with the human model, and vice-versa, achieving ≈97.80% accuracy in both cases, showing that the model is not overfit. The SVM models were validated with data from rat (Rattus norvegicus), pig (Sus scrofa) and fruit fly (Drosophila melanogaster), and obtained more than 84.00% accuracy in all these organisms. Our results also showed that 81.2% of human pseudogenes and 91.7% of mouse pseudogenes were classified as non-coding. Moreover, our method was capable of re-annotating two uncharacterized sequences of Swiss-Prot database with high probability of being lncRNAs. Finally, in order to use the method to annotate transcripts derived from RNA-seq, previously identified lncRNAs of human, gorilla (Gorilla gorilla) and rhesus macaque (Macaca mulatta) were analyzed, having successfully classified 98.62%, 80.8% and 91.9%, respectively. CONCLUSIONS: The SVM method proposed in this work presents high performance to distinguish lncRNAs from PCTs, as shown in the results. To build the model, besides using features known in the literature regarding ORFs, we used PCA to identify features among nucleotide pattern frequencies that contribute the most in distinguishing lncRNAs from PCTs, in reference data sets. Interestingly, models created with two evolutionary distant species could distinguish lncRNAs of even more distant species.

SnoReport 2.0: new features and a refined Support Vector Machine to improve snoRNA identification.

BACKGROUND: snoReport uses RNA secondary structure prediction combined with machine learning as the basis to identify the two main classes of small nucleolar RNAs, the box H/ACA snoRNAs and the box C/D snoRNAs. Here, we present snoReport 2.0, which substantially improves and extends in the original method by: extracting new features for both box C/D and H/ACA box snoRNAs; developing a more sophisticated technique in the SVM training phase with recent data from vertebrate organisms and a careful choice of the SVM parameters C and γ; and using updated versions of tools and databases used for the construction of the original version of snoReport. To validate the new version and to demonstrate its improved performance, we tested snoReport 2.0 in different organisms. RESULTS: Results of the training and test phases of boxes H/ACA and C/D snoRNAs, in both versions of snoReport, are discussed. Validation on real data was performed to evaluate the predictions of snoReport 2.0. Our program was applied to a set of previously annotated sequences, some of them experimentally confirmed, of humans, nematodes, drosophilids, platypus, chickens and leishmania. We significantly improved the predictions for vertebrates, since the training phase used information of these organisms, but H/ACA box snoRNAs identification was improved for the other ones. CONCLUSION: We presented snoReport 2.0, to predict H/ACA box and C/D box snoRNAs, an efficient method to find true positives and avoid false positives in vertebrate organisms. H/ACA box snoRNA classifier showed an F-score of 93 % (an improvement of 10 % regarding the previous version), while C/D box snoRNA classifier, an F-Score of 94 % (improvement of 14 %). Besides, both classifiers exhibited performance measures above 90 %. These results show that snoReport 2.0 avoid false positives and false negatives, allowing to predict snoRNAs with high quality. In the validation phase, snoReport 2.0 predicted 67.43 % of vertebrate organisms for both classes. For Nematodes and Drosophilids, 69 % and 76.67 %, for H/ACA box snoRNAs were predicted, respectively, showing that snoReport 2.0 is good to identify snoRNAs in vertebrates and also H/ACA box snoRNAs in invertebrates organisms.

Odorant recognition using biological responses recorded in olfactory bulb of rats.

In this study we applied pattern recognition (PR) techniques to extract odorant information from local field potential (LFP) signals recorded in the olfactory bulb (OB) of rats subjected to different odorant stimuli. We claim that LFP signals registered on the OB, the first stage of olfactory processing, are stimulus specific in animals with normal sensory experience, and that these patterns correspond to the neural substrate likely required for perceptual discrimination. Thus, these signals can be used as input to an artificial odorant classification system with great success. In this paper we have designed and compared the performance of several configurations of artificial olfaction systems (AOS) based on the combination of four feature extraction (FE) methods (Principal Component Analysis (PCA), Fisher Transformation (FT), Sammon NonLinear Map (NLM) and Wavelet Transform (WT)), and three PR techniques (Linear Discriminant Analysis (LDA), Multilayer Perceptron (MLP) and Support Vector Machine (SVM)), when four different stimuli are presented to rats. The best results were reached when PCA extraction followed by SVM as classifier were used, obtaining a classification accuracy of over 95% for all four stimuli.

Clinician's road map to wavelet EEG as an Alzheimer's disease biomarker.

Alzheimer's disease (AD) is considered the main cause of dementia in Western countries. Consequently, there is a need for an accurate, universal, specific and cost-effective biomarker for early AD diagnosis, to follow disease progression and therapy response. This article describes a new diagnostic approach to quantitative electroencephalogram (QEEG) diagnosis of mild and moderate AD. The data set used in this study was composed of EEG signals recorded from 2 groups: (S1) 74 normal subjects, 33 females and 41 males (mean age 67 years, standard deviation = 8) and (S2) 88 probable AD patients (NINCDS-ADRDA criteria), 55 females and 33 males (mean age 74.7 years, standard deviation = 7.8) with mild to moderate symptoms (DSM-IV-TR). Attention is given to sample size and the use of state of the art open source tools (LetsWave and WEKA) to process the EEG data. This innovative technique consists in associating Morlet wavelet filter with a support vector machine technique. A total of 111 EEG features (attributes) were obtained for 162 probands. The results were accuracy of 92.72% and area under the curve of 0.92 (percentage split test). Most important, comparing a single patient versus the total data set resulted in accuracy of 84.56% (leave-one-patient-out test). Particular emphasis was on clinical diagnosis and feasibility of implementation of this low-cost procedure, because programming knowledge is not required. Consequently, this new method can be useful to support AD diagnosis in resource-limited settings.