An interpretable block-attention network for identifying regulatory feature interactions.

The importance of regulatory features in health and disease is increasing, making it crucial to identify the hallmarks of these features. Self-attention networks (SAN) have given rise to numerous models for the prediction of complex phenomena. But the potential of SANs in biological models was limited because of high memory requirement proportional to input token length and lack of interpretability of self-attention scores. To overcome these constraints, we propose a deep learning model named Interpretable Self-Attention Network for REGulatory interactions (ISANREG) that combines both block self-attention and attention-attribution mechanisms. This model predicts transcription factor-bound motif instances and DNA-mediated TF-TF interactions using self-attention attribution scores derived from the network, overcoming the limitations of previous deep learning models. ISANREG will serve as a framework for other biological models in interpreting the contribution of the input with single-nucleotide resolution.

TIANA: transcription factors cooperativity inference analysis with neural attention.

BACKGROUND: Growing evidence suggests that distal regulatory elements are essential for cellular function and states. The sequences within these distal elements, especially motifs for transcription factor binding, provide critical information about the underlying regulatory programs. However, cooperativities between transcription factors that recognize these motifs are nonlinear and multiplexed, rendering traditional modeling methods insufficient to capture the underlying mechanisms. Recent development of attention mechanism, which exhibit superior performance in capturing dependencies across input sequences, makes them well-suited to uncover and decipher intricate dependencies between regulatory elements. RESULT: We present Transcription factors cooperativity Inference Analysis with Neural Attention (TIANA), a deep learning framework that focuses on interpretability. In this study, we demonstrated that TIANA could discover biologically relevant insights into co-occurring pairs of transcription factor motifs. Compared with existing tools, TIANA showed superior interpretability and robust performance in identifying putative transcription factor cooperativities from co-occurring motifs. CONCLUSION: Our results suggest that TIANA can be an effective tool to decipher transcription factor cooperativities from distal sequence data. TIANA can be accessed through: https://github.com/rzzli/TIANA .

Clinical domain knowledge-derived template improves post hoc AI explanations in pneumothorax classification.

OBJECTIVE: Pneumothorax is an acute thoracic disease caused by abnormal air collection between the lungs and chest wall. Recently, artificial intelligence (AI), especially deep learning (DL), has been increasingly employed for automating the diagnostic process of pneumothorax. To address the opaqueness often associated with DL models, explainable artificial intelligence (XAI) methods have been introduced to outline regions related to pneumothorax. However, these explanations sometimes diverge from actual lesion areas, highlighting the need for further improvement. METHOD: We propose a template-guided approach to incorporate the clinical knowledge of pneumothorax into model explanations generated by XAI methods, thereby enhancing the quality of the explanations. Utilizing one lesion delineation created by radiologists, our approach first generates a template that represents potential areas of pneumothorax occurrence. This template is then superimposed on model explanations to filter out extraneous explanations that fall outside the template's boundaries. To validate its efficacy, we carried out a comparative analysis of three XAI methods (Saliency Map, Grad-CAM, and Integrated Gradients) with and without our template guidance when explaining two DL models (VGG-19 and ResNet-50) in two real-world datasets (SIIM-ACR and ChestX-Det). RESULTS: The proposed approach consistently improved baseline XAI methods across twelve benchmark scenarios built on three XAI methods, two DL models, and two datasets. The average incremental percentages, calculated by the performance improvements over the baseline performance, were 97.8% in Intersection over Union (IoU) and 94.1% in Dice Similarity Coefficient (DSC) when comparing model explanations and ground-truth lesion areas. We further visualized baseline and template-guided model explanations on radiographs to showcase the performance of our approach. CONCLUSIONS: In the context of pneumothorax diagnoses, we proposed a template-guided approach for improving model explanations. Our approach not only aligns model explanations more closely with clinical insights but also exhibits extensibility to other thoracic diseases. We anticipate that our template guidance will forge a novel approach to elucidating AI models by integrating clinical domain expertise.

An Explainable Deep Learning Approach for Stress Detection in Wearable Sensor Measurements.

Stress has various impacts on the health of human beings. Recent success in wearable sensor development, combined with advancements in deep learning to automatically detect features from raw data, opens several interesting applications related to detecting emotional states. Being able to accurately detect stress-related emotional arousal in an acute setting can positively impact the imminent health status of humans, i.e., through avoiding dangerous locations in an urban traffic setting. This work proposes an explainable deep learning methodology for the automatic detection of stress in physiological sensor data, recorded through a non-invasive wearable sensor device, the Empatica E4 wristband. We propose a Long-Short Term-Memory (LSTM) network, extended through a Deep Generative Ensemble of conditional GANs (LSTM DGE), to deal with the low data regime of sparsely labeled sensor measurements. As explainability is often a main concern of deep learning models, we leverage Integrated Gradients (IG) to highlight the most essential features used by the model for prediction and to compare the results to state-of-the-art expert-based stress-detection methodologies in terms of precision, recall, and interpretability. The results show that our LSTM DGE outperforms the state-of-the-art algorithm by 3 percentage points in terms of recall, and 7.18 percentage points in terms of precision. More importantly, through the use of Integrated Gradients as a layer of explainability, we show that there is a strong overlap between model-derived stress features for electrodermal activity and existing literature, which current state-of-the-art stress detection systems in medical research and psychology are based on.

Building a trustworthy AI differential diagnosis application for Crohn's disease and intestinal tuberculosis.

BACKGROUND: Differentiating between Crohn's disease (CD) and intestinal tuberculosis (ITB) with endoscopy is challenging. We aim to perform more accurate endoscopic diagnosis between CD and ITB by building a trustworthy AI differential diagnosis application. METHODS: A total of 1271 electronic health record (EHR) patients who had undergone colonoscopies at Peking Union Medical College Hospital (PUMCH) and were clinically diagnosed with CD (n = 875) or ITB (n = 396) were used in this study. We build a workflow to make diagnoses with EHRs and mine differential diagnosis features; this involves finetuning the pretrained language models, distilling them into a light and efficient TextCNN model, interpreting the neural network and selecting differential attribution features, and then adopting manual feature checking and carrying out debias training. RESULTS: The accuracy of debiased TextCNN on differential diagnosis between CD and ITB is 0.83 (CR F1: 0.87, ITB F1: 0.77), which is the best among the baselines. On the noisy validation set, its accuracy was 0.70 (CR F1: 0.87, ITB: 0.69), which was significantly higher than that of models without debias. We also find that the debiased model more easily mines the diagnostically significant features. The debiased TextCNN unearthed 39 diagnostic features in the form of phrases, 17 of which were key diagnostic features recognized by the guidelines. CONCLUSION: We build a trustworthy AI differential diagnosis application for differentiating between CD and ITB focusing on accuracy, interpretability and robustness. The classifiers perform well, and the features which had statistical significance were in agreement with clinical guidelines.

Integrated Gradient-Based Continuous Wavelet Transform for Bearing Fault Diagnosis.

Bearing fault diagnosis is important to ensure safe operation and reduce loss for most rotating machinery. In recent years, deep learning (DL) has been widely used for bearing fault diagnosis and has achieved excellent results. Continuous wavelet transform (CWT), which can convert original sensor data to time-frequency images, is often used to preprocess vibration data for the DL model. However, in time-frequency images, some frequency components may be important, and some may be unimportant for DL models for fault diagnosis. So, how to choose a frequency range of important frequency components is needed for CWT. In this paper, an Integrated Gradient-based continuous wavelet transform (IG-CWT) method is proposed to address this issue. Through IG-CWT, the important frequency components and the component frequency range can be detected and used for data preprocessing. To verify our method, experiments are conducted on four famous bearing datasets using 3 DL models, separately, and compared with CWT, and the results are compared with the original CWT. The comparisons show that the proposed IG-CWT can achieve higher fault diagnosis accuracy.

Multi-resBind: a residual network-based multi-label classifier for in vivo RNA binding prediction and preference visualization.

BACKGROUND: Protein-RNA interactions play key roles in many processes regulating gene expression. To understand the underlying binding preference, ultraviolet cross-linking and immunoprecipitation (CLIP)-based methods have been used to identify the binding sites for hundreds of RNA-binding proteins (RBPs) in vivo. Using these large-scale experimental data to infer RNA binding preference and predict missing binding sites has become a great challenge. Some existing deep-learning models have demonstrated high prediction accuracy for individual RBPs. However, it remains difficult to avoid significant bias due to the experimental protocol. The DeepRiPe method was recently developed to solve this problem via introducing multi-task or multi-label learning into this field. However, this method has not reached an ideal level of prediction power due to the weak neural network architecture. RESULTS: Compared to the DeepRiPe approach, our Multi-resBind method demonstrated substantial improvements using the same large-scale PAR-CLIP dataset with respect to an increase in the area under the receiver operating characteristic curve and average precision. We conducted extensive experiments to evaluate the impact of various types of input data on the final prediction accuracy. The same approach was used to evaluate the effect of loss functions. Finally, a modified integrated gradient was employed to generate attribution maps. The patterns disentangled from relative contributions according to context offer biological insights into the underlying mechanism of protein-RNA interactions. CONCLUSIONS: Here, we propose Multi-resBind as a new multi-label deep-learning approach to infer protein-RNA binding preferences and predict novel interactions. The results clearly demonstrate that Multi-resBind is a promising tool to predict unknown binding sites in vivo and gain biology insights into why the neural network makes a given prediction.

Lifelog Data-Based Prediction Model of Digital Health Care App Customer Churn: Retrospective Observational Study.

BACKGROUND: Customer churn is the rate at which customers stop doing business with an entity. In the field of digital health care, user churn prediction is important not only in terms of company revenue but also for improving the health of users. Churn prediction has been previously studied, but most studies applied time-invariant model structures and used structured data. However, additional unstructured data have become available; therefore, it has become essential to process daily time-series log data for churn predictions. OBJECTIVE: We aimed to apply a recurrent neural network structure to accept time-series patterns using lifelog data and text message data to predict the churn of digital health care users. METHODS: This study was based on the use data of a digital health care app that provides interactive messages with human coaches regarding food, exercise, and weight logs. Among the users in Korea who enrolled between January 1, 2017 and January 1, 2019, we defined churn users according to the following criteria: users who received a refund before the paid program ended and users who received a refund 7 days after the trial period. We used long short-term memory with a masking layer to receive sequence data with different lengths. We also performed topic modeling to vectorize text messages. To interpret the contributions of each variable to model predictions, we used integrated gradients, which is an attribution method. RESULTS: A total of 1868 eligible users were included in this study. The final performance of churn prediction was an F1 score of 0.89; that score decreased by 0.12 when the data of the final week were excluded (F1 score 0.77). Additionally, when text data were included, the mean predicted performance increased by approximately 0.085 at every time point. Steps per day had the largest contribution (0.1085). Among the topic variables, poor habits (eg, drinking alcohol, overeating, and late-night eating) showed the largest contribution (0.0875). CONCLUSIONS: The model with a recurrent neural network architecture that used log data and message data demonstrated high performance for churn classification. Additionally, the analysis of the contribution of the variables is expected to help identify signs of user churn in advance and improve the adherence in digital health care.

An interpretable generative multimodal neuroimaging-genomics framework for decoding Alzheimer's disease.

Alzheimer's disease (AD) is the most prevalent form of dementia, affecting millions worldwide with a progressive decline in cognitive abilities. The AD continuum encompasses a prodormal stage known as Mild Cognitive Impairment (MCI), where patients may either progress to AD (MCIc) or remain stable (MCInc). Understanding the underlying mechanisms of AD requires complementary analysis derived from different data sources, leading to the development of multimodal deep learning models. In this study, we leveraged structural and functional Magnetic Resonance Imaging (sMRI/fMRI) to investigate the disease-induced grey matter and functional network connectivity changes. Moreover, considering AD's strong genetic component, we introduce Single Nucleotide Polymorphisms (SNPs) as a third channel. Given such diverse inputs, missing one or more modalities is a typical concern of multimodal methods. We hence propose a novel deep learning based classification framework where generative module employing Cycle Generative Adversarial Networks (cGAN) was adopted to impute missing data within the latent space. Additionally, we adopted an Explainable Artificial Intelligence (XAI) method, Integrated Gradients (IG), to extract input features relevance, enhancing our understanding of the learned representations. Two critical tasks were addressed: AD detection and MCI conversion prediction. Experimental results showed that our framework was able to reach the state-of-the-art in the classification of CN vs AD reaching an average test accuracy of 0.926 ± 0.02. For the MCInc vs MCIc task, we achieved an average prediction accuracy of 0.711 ± 0.01 using the pre-trained model for CN and AD. The interpretability analysis revealed that the classification performance was led by significant grey matter modulations in cortical and subcortical brain areas well known for their association with AD. Moreover, impairments in sensory-motor and visual resting state network connectivity along the disease continuum, as well as mutations in SNPs defining biological processes linked to amyloid-beta and cholesterol formation clearance and regulation, were identified as contributors to the achieved performance. Overall, our integrative deep learning approach shows promise for AD detection and MCI prediction, while shading light on important biological insights.

A clinically actionable and explainable real-time risk assessment framework for stroke-associated pneumonia.

The current medical practice is more responsive rather than proactive, despite the widely recognized value of early disease detection, including improving the quality of care and reducing medical costs. One of the cornerstones of early disease detection is clinically actionable predictions, where predictions are expected to be accurate, stable, real-time and interpretable. As an example, we used stroke-associated pneumonia (SAP), setting up a transformer-encoder-based model that analyzes highly heterogeneous electronic health records in real-time. The model was proven accurate and stable on an independent test set. In addition, it issued at least one warning for 98.6 % of SAP patients, and on average, its alerts were ahead of physician diagnoses by 2.71 days. We applied Integrated Gradient to glean the model's reasoning process. Supplementing the risk scores, the model highlighted critical historical events on patients' trajectories, which were shown to have high clinical relevance.

MeSH2Matrix: combining MeSH keywords and machine learning for biomedical relation classification based on PubMed.

Biomedical relation classification has been significantly improved by the application of advanced machine learning techniques on the raw texts of scholarly publications. Despite this improvement, the reliance on large chunks of raw text makes these algorithms suffer in terms of generalization, precision, and reliability. The use of the distinctive characteristics of bibliographic metadata can prove effective in achieving better performance for this challenging task. In this research paper, we introduce an approach for biomedical relation classification using the qualifiers of co-occurring Medical Subject Headings (MeSH). First of all, we introduce MeSH2Matrix, our dataset consisting of 46,469 biomedical relations curated from PubMed publications using our approach. Our dataset includes a matrix that maps associations between the qualifiers of subject MeSH keywords and those of object MeSH keywords. It also specifies the corresponding Wikidata relation type and the superclass of semantic relations for each relation. Using MeSH2Matrix, we build and train three machine learning models (Support Vector Machine [SVM], a dense model [D-Model], and a convolutional neural network [C-Net]) to evaluate the efficiency of our approach for biomedical relation classification. Our best model achieves an accuracy of 70.78% for 195 classes and 83.09% for five superclasses. Finally, we provide confusion matrix and extensive feature analyses to better examine the relationship between the MeSH qualifiers and the biomedical relations being classified. Our results will hopefully shed light on developing better algorithms for biomedical ontology classification based on the MeSH keywords of PubMed publications. For reproducibility purposes, MeSH2Matrix, as well as all our source codes, are made publicly accessible at https://github.com/SisonkeBiotik-Africa/MeSH2Matrix .

Explainable AI to Facilitate Understanding of Neural Network-Based Metabolite Profiling Using NMR Spectroscopy.

Neural networks (NNs) are emerging as a rapid and scalable method for quantifying metabolites directly from nuclear magnetic resonance (NMR) spectra, but the nonlinear nature of NNs precludes understanding of how a model makes predictions. This study implements an explainable artificial intelligence algorithm called integrated gradients (IG) to elucidate which regions of input spectra are the most important for the quantification of specific analytes. The approach is first validated in simulated mixture spectra of eight aqueous metabolites and then investigated in experimentally acquired lipid spectra of a reference standard mixture and a murine hepatic extract. The IG method revealed that, like a human spectroscopist, NNs recognize and quantify analytes based on an analyte's respective resonance line-shapes, amplitudes, and frequencies. NNs can compensate for peak overlap and prioritize specific resonances most important for concentration determination. Further, we show how modifying a NN training dataset can affect how a model makes decisions, and we provide examples of how this approach can be used to de-bug issues with model performance. Overall, results show that the IG technique facilitates a visual and quantitative understanding of how model inputs relate to model outputs, potentially making NNs a more attractive option for targeted and automated NMR-based metabolomics.

Interpretable deep learning reveals the role of an E-box motif in suppressing somatic hypermutation of AGCT motifs within human immunoglobulin variable regions.

Introduction: Somatic hypermutation (SHM) of immunoglobulin variable (V) regions by activation induced deaminase (AID) is essential for robust, long-term humoral immunity against pathogen and vaccine antigens. AID mutates cytosines preferentially within WRCH motifs (where W=A or T, R=A or G and H=A, C or T). However, it has been consistently observed that the mutability of WRCH motifs varies substantially, with large variations in mutation frequency even between multiple occurrences of the same motif within a single V region. This has led to the notion that the immediate sequence context of WRCH motifs contributes to mutability. Recent studies have highlighted the potential role of local DNA sequence features in promoting mutagenesis of AGCT, a commonly mutated WRCH motif. Intriguingly, AGCT motifs closer to 5' ends of V regions, within the framework 1 (FW1) sub-region1, mutate less frequently, suggesting an SHM-suppressing sequence context. Methods: Here, we systematically examined the basis of AGCT positional biases in human SHM datasets with DeepSHM, a machine-learning model designed to predict SHM patterns. This was combined with integrated gradients, an interpretability method, to interrogate the basis of DeepSHM predictions. Results: DeepSHM predicted the observed positional differences in mutation frequencies at AGCT motifs with high accuracy. For the conserved, lowly mutating AGCT motifs in FW1, integrated gradients predicted a large negative contribution of 5'C and 3'G flanking residues, suggesting that a CAGCTG context in this location was suppressive for SHM. CAGCTG is the recognition motif for E-box transcription factors, including E2A, which has been implicated in SHM. Indeed, we found a strong, inverse relationship between E-box motif fidelity and mutation frequency. Moreover, E2A was found to associate with the V region locale in two human B cell lines. Finally, analysis of human SHM datasets revealed that naturally occurring mutations in the 3'G flanking residues, which effectively ablate the E-box motif, were associated with a significantly increased rate of AGCT mutation. Discussion: Our results suggest an antagonistic relationship between mutation frequency and the binding of E-box factors like E2A at specific AGCT motif contexts and, therefore, highlight a new, suppressive mechanism regulating local SHM patterns in human V regions.

Identification of B cell subsets based on antigen receptor sequences using deep learning.

B cell receptors (BCRs) denote antigen specificity, while corresponding cell subsets indicate B cell functionality. Since each B cell uniquely encodes this combination, physical isolation and subsequent processing of individual B cells become indispensable to identify both attributes. However, this approach accompanies high costs and inevitable information loss, hindering high-throughput investigation of B cell populations. Here, we present BCR-SORT, a deep learning model that predicts cell subsets from their corresponding BCR sequences by leveraging B cell activation and maturation signatures encoded within BCR sequences. Subsequently, BCR-SORT is demonstrated to improve reconstruction of BCR phylogenetic trees, and reproduce results consistent with those verified using physical isolation-based methods or prior knowledge. Notably, when applied to BCR sequences from COVID-19 vaccine recipients, it revealed inter-individual heterogeneity of evolutionary trajectories towards Omicron-binding memory B cells. Overall, BCR-SORT offers great potential to improve our understanding of B cell responses.

Different Recognition of Protein Features Depending on Deep Learning Models: A Case Study of Aromatic Decarboxylase UbiD.

The number of unannotated protein sequences is explosively increasing due to genome sequence technology. A more comprehensive understanding of protein functions for protein annotation requires the discovery of new features that cannot be captured from conventional methods. Deep learning can extract important features from input data and predict protein functions based on the features. Here, protein feature vectors generated by 3 deep learning models are analyzed using Integrated Gradients to explore important features of amino acid sites. As a case study, prediction and feature extraction models for UbiD enzymes were built using these models. The important amino acid residues extracted from the models were different from secondary structures, conserved regions and active sites of known UbiD information. Interestingly, the different amino acid residues within UbiD sequences were regarded as important factors depending on the type of models and sequences. The Transformer models focused on more specific regions than the other models. These results suggest that each deep learning model understands protein features with different aspects from existing knowledge and has the potential to discover new laws of protein functions. This study will help to extract new protein features for the other protein annotations.

Spatial transformation of multi-omics data unlocks novel insights into cancer biology.

The application of next-generation sequencing (NGS) has transformed cancer research. As costs have decreased, NGS has increasingly been applied to generate multiple layers of molecular data from the same samples, covering genomics, transcriptomics, and methylomics. Integrating these types of multi-omics data in a combined analysis is now becoming a common issue with no obvious solution, often handled on an ad hoc basis, with multi-omics data arriving in a tabular format and analyzed using computationally intensive statistical methods. These methods particularly ignore the spatial orientation of the genome and often apply stringent p-value corrections that likely result in the loss of true positive associations. Here, we present GENIUS (GEnome traNsformatIon and spatial representation of mUltiomicS data), a framework for integrating multi-omics data using deep learning models developed for advanced image analysis. The GENIUS framework is able to transform multi-omics data into images with genes displayed as spatially connected pixels and successfully extract relevant information with respect to the desired output. We demonstrate the utility of GENIUS by applying the framework to multi-omics datasets from the Cancer Genome Atlas. Our results are focused on predicting the development of metastatic cancer from primary tumors, and demonstrate how through model inference, we are able to extract the genes which are driving the model prediction and are likely associated with metastatic disease progression. We anticipate our framework to be a starting point and strong proof of concept for multi-omics data transformation and analysis without the need for statistical correction.

Compensated Integrated Gradients for Reliable Explanation of Electroencephalogram Signal Classification.

The integrated gradients (IG) method is widely used to evaluate the extent to which each input feature contributes to the classification using a deep learning model because it theoretically satisfies the desired properties to fairly attribute the contributions to the classification. However, this approach requires an appropriate baseline to do so. In this study, we propose a compensated IG method that does not require a baseline, which compensates the contributions calculated using the IG method at an arbitrary baseline by using an example of the Shapley sampling value. We prove that the proposed approach can compute the contributions to the classification results reliably if the processes of each input feature in a classifier are independent of one another and the parameterization of each process is identical, as in shared weights in convolutional neural networks. Using three datasets on electroencephalogram recordings, we experimentally demonstrate that the contributions obtained by the proposed compensated IG method are more reliable than those obtained using the original IG method and that its computational complexity is much lower than that of the Shapley sampling method.

Assessing comparative importance of DNA sequence and epigenetic modifications on gene expression using a deep convolutional neural network.

Gene expression is regulated at both transcriptional and post-transcriptional levels. DNA sequence and epigenetic modifications are key factors which regulate gene transcription. Understanding their complex interactions and their respective contributions to gene expression regulation remains a challenge in biological studies. We have developed iSEGnet, a framework of deep convolutional neural network to predict mRNA abundance using the information on DNA sequences as well as epigenetic modifications within genes and their cis-regulatory regions. We demonstrate that our framework outperforms other machine learning models in terms of predicting mRNA abundance using transcriptional and epigenetic profiles from six distinct cell lines/types chosen from the ENCODE. The analysis from the learned models also reveals that specific regions around promotors and transcription termination sites are most important for gene expression regulation. Using the method of Integrated Gradients, we identify narrow segments in these regions which are most likely to impact gene expression for a specific epigenetic modification. We further show that these identified segments are enriched in known active regulatory regions by comparing the transcription factor binding sites obtained via ChIP-seq. Moreover, we demonstrate how iSEGnet can uncover potential transcription factors that have regulatory functions in cancer using two cancer multi-omics data.

Data-driven electrophysiological feature based on deep learning to detect epileptic seizures.

Objective. To identify a new electrophysiological feature characterising the epileptic seizures, which is commonly observed in different types of epilepsy.Methods. We recorded the intracranial electroencephalogram (iEEG) of 21 patients (12 women and 9 men) with multiple types of refractory epilepsy. The raw iEEG signals of the early phase of epileptic seizures and interictal states were classified by a convolutional neural network (Epi-Net). For comparison, the same signals were classified by a support vector machine (SVM) using the spectral power and phase-amplitude coupling. The features learned by Epi-Net were derived by a modified integrated gradients method. We considered the product of powers multiplied by the relative contribution of each frequency amplitude as a data-driven epileptogenicity index (d-EI). We compared the d-EI and other conventional features in terms of accuracy to detect the epileptic seizures. Finally, we compared the d-EI among the electrodes to evaluate its relationship with the resected area and the Engel classification.Results. Epi-Net successfully identified the epileptic seizures, with an area under the receiver operating characteristic curve of 0.944 ± 0.067, which was significantly larger than that of the SVM (0.808 ± 0.253,n =21;p =0.025). The learned iEEG signals were characterised by increased powers of 17-92 Hz and >180 Hz in addition to decreased powers of other frequencies. The proposed d-EI detected them with better accuracy than the other iEEG features. Moreover, the surgical resection of areas with a larger increase in d-EI was observed for all nine patients with Engel class ⩽1, but not for the 4 of 12 patients with Engel class >1, demonstrating the significant association with seizure outcomes.Significance.We derived an iEEG feature from the trained Epi-Net, which identified the epileptic seizures with improved accuracy and might contribute to identification of the epileptogenic zone.

SpliceViNCI: Visualizing the splicing of non-canonical introns through recurrent neural networks.

Most of the current computational models for splice junction prediction are based on the identification of canonical splice junctions. However, it is observed that the junctions lacking the consensus dimers GT and AG also undergo splicing. Identification of such splice junctions, called the non-canonical splice junctions, is also essential for a comprehensive understanding of the splicing phenomenon. This work focuses on the identification of non-canonical splice junctions through the application of a bidirectional long short-term memory (BLSTM) network. Furthermore, we apply a back-propagation-based (integrated gradient) and a perturbation-based (occlusion) visualization techniques to extract the non-canonical splicing features learned by the model. The features obtained are validated with the existing knowledge from the literature. Integrated gradient extracts features that comprise contiguous nucleotides, whereas occlusion extracts features that are individual nucleotides distributed across the sequence.