Md. Rezaul Karim

Md. Rezaul Karim, Tanhim Islam, Oya Beyan et al.

Artificial intelligence (AI) systems utilizing deep neural networks (DNNs) and machine learning (ML) algorithms are widely used for solving important problems in bioinformatics, biomedical informatics, and precision medicine. However, complex DNNs or ML models, which are often perceived as opaque and black-box, can make it difficult to understand the reasoning behind their decisions. This lack of transparency can be a challenge for both end-users and decision-makers, as well as AI developers. Additionally, in sensitive areas like healthcare, explainability and accountability are not only desirable but also legally required for AI systems that can have a significant impact on human lives. Fairness is another growing concern, as algorithmic decisions should not show bias or discrimination towards certain groups or individuals based on sensitive attributes. Explainable artificial intelligence (XAI) aims to overcome the opaqueness of black-box models and provide transparency in how AI systems make decisions. Interpretable ML models can explain how they make predictions and the factors that influence their outcomes. However, most state-of-the-art interpretable ML methods are domain-agnostic and evolved from fields like computer vision, automated reasoning, or statistics, making direct application to bioinformatics problems challenging without customization and domain-specific adaptation. In this paper, we discuss the importance of explainability in the context of bioinformatics, provide an overview of model-specific and model-agnostic interpretable ML methods and tools, and outline their potential caveats and drawbacks. Besides, we discuss how to customize existing interpretable ML methods for bioinformatics problems. Nevertheless, we demonstrate how XAI methods can improve transparency through case studies in bioimaging, cancer genomics, and text mining.

Md. Rezaul Karim, Sumon Kanti Dey, Tanhim Islam et al.

Numerous machine learning (ML) and deep learning (DL)-based approaches have been proposed to utilize textual data from social media for anti-social behavior analysis like cyberbullying, fake news detection, and identification of hate speech mainly for highly-resourced languages such as English. However, despite having a lot of diversity and millions of native speakers, some languages like Bengali are under-resourced, which is due to a lack of computational resources for natural language processing (NLP). Similar to other languages, Bengali social media contents also include images along with texts (e.g., multimodal memes are posted by embedding short texts into images on Facebook). Therefore, only the textual data is not enough to judge them since images might give extra context to make a proper judgement. This paper is about hate speech detection from multimodal Bengali memes and texts. We prepared the only multimodal hate speech dataset for-a-kind of problem for Bengali, which we use to train state-of-the-art neural architectures (e.g., Bi-LSTM/Conv-LSTM with word embeddings, ConvNets + pre-trained language models, e.g., monolingual Bangla BERT, multilingual BERT-cased/uncased, and XLM-RoBERTa) to jointly analyze textual and visual information for hate speech detection. Conv-LSTM and XLM-RoBERTa models performed best for texts, yielding F1 scores of 0.78 and 0.82, respectively. As of memes, ResNet-152 and DenseNet-161 models yield F1 scores of 0.78 and 0.79, respectively. As for multimodal fusion, XLM-RoBERTa + DenseNet-161 performed the best, yielding an F1 score of 0.83. Our study suggests that text modality is most useful for hate speech detection, while memes are moderately useful.

Md. Rezaul Karim, Md. Shajalal, Alex Graß et al.

Deep neural networks (DNNs) have been shown to outperform traditional machine learning algorithms in a broad variety of application domains due to their effectiveness in modeling complex problems and handling high-dimensional datasets. Many real-life datasets, however, are of increasingly high dimensionality, where a large number of features may be irrelevant for both supervised and unsupervised learning tasks. The inclusion of such features would not only introduce unwanted noise but also increase computational complexity. Furthermore, due to high non-linearity and dependency among a large number of features, DNN models tend to be unavoidably opaque and perceived as black-box methods because of their not well-understood internal functioning. Their algorithmic complexity is often simply beyond the capacities of humans to understand the interplay among myriads of hyperparameters. A well-interpretable model can identify statistically significant features and explain the way they affect the model's outcome. In this paper, we propose an efficient method to improve the interpretability of black-box models for classification tasks in the case of high-dimensional datasets. First, we train a black-box model on a high-dimensional dataset to learn the embeddings on which the classification is performed. To decompose the inner working principles of the black-box model and to identify top-k important features, we employ different probing and perturbing techniques. We then approximate the behavior of the black-box model by means of an interpretable surrogate model on the top-k feature space. Finally, we derive decision rules and local explanations from the surrogate model to explain individual decisions. Our approach outperforms state-of-the-art methods like TabNet and XGboost when tested on different datasets with varying dimensionality between 50 and 20,000 w.r.t metrics and explainability.

Md. Rezaul Karim, Hussain Ali, Prinon Das et al.

Structured and unstructured data and facts about drugs, genes, protein, viruses, and their mechanism are spread across a huge number of scientific articles. These articles are a large-scale knowledge source and can have a huge impact on disseminating knowledge about the mechanisms of certain biological processes. A knowledge graph (KG) can be constructed by integrating such facts and data and be used for data integration, exploration, and federated queries. However, exploration and querying large-scale KGs is tedious for certain groups of users due to a lack of knowledge about underlying data assets or semantic technologies. A question-answering (QA) system allows the answer of natural language questions over KGs automatically using triples contained in a KG. Recently, the use and adaption of digital assistants are getting wider owing to their capability at enabling users to voice commands to control smart systems or devices. This paper is about using Amazon Alexa's voice-enabled interface for QA over KGs. As a proof-of-concept, we use the well-known DisgeNET KG, which contains knowledge covering 1.13 million gene-disease associations between 21,671 genes and 30,170 diseases, disorders, and clinical or abnormal human phenotypes. Our study shows how Alex could be of help to find facts about certain biological entities from large-scale knowledge bases.

Md. Rezaul Karim, Lina Molinas Comet, Md Shajalal et al.

Domain experts often rely on most recent knowledge for apprehending and disseminating specific biological processes that help them design strategies for developing prevention and therapeutic decision-making in various disease scenarios. A challenging scenarios for artificial intelligence (AI) is using biomedical data (e.g., texts, imaging, omics, and clinical) to provide diagnosis and treatment recommendations for cancerous conditions.~Data and knowledge about biomedical entities like cancer, drugs, genes, proteins, and their mechanism is spread across structured (knowledge bases (KBs)) and unstructured (e.g., scientific articles) sources. A large-scale knowledge graph (KG) can be constructed by integrating and extracting facts about semantically interrelated entities and relations. Such a KG not only allows exploration and question answering (QA) but also enables domain experts to deduce new knowledge. However, exploring and querying large-scale KGs is tedious for non-domain users due to their lack of understanding of the data assets and semantic technologies. In this paper, we develop a domain KG to leverage cancer-specific biomarker discovery and interactive QA. For this, we constructed a domain ontology called OncoNet Ontology (ONO), which enables semantic reasoning for validating gene-disease (different types of cancer) relations. The KG is further enriched by harmonizing the ONO, metadata, controlled vocabularies, and biomedical concepts from scientific articles by employing BioBERT- and SciBERT-based information extractors. Further, since the biomedical domain is evolving, where new findings often replace old ones, without having access to up-to-date scientific findings, there is a high chance an AI system exhibits concept drift while providing diagnosis and treatment. Therefore, we fine-tune the KG using large language models (LLMs) based on more recent articles and KBs.

Md. Rezaul Karim, Lina Molinas Comet, Oya Beyan et al.

Structured and unstructured data and facts about drugs, genes, protein, viruses, and their mechanism are spread across a huge number of scientific articles. These articles are a large-scale knowledge source and can have a huge impact on disseminating knowledge about the mechanisms of certain biological processes. A domain-specific knowledge graph~(KG) is an explicit conceptualization of a specific subject-matter domain represented w.r.t semantically interrelated entities and relations. A KG can be constructed by integrating such facts and data and be used for data integration, exploration, and federated queries. However, exploration and querying large-scale KGs is tedious for certain groups of users due to a lack of knowledge about underlying data assets or semantic technologies. Such a KG will not only allow deducing new knowledge and question answering(QA) but also allows domain experts to explore. Since cross-disciplinary explanations are important for accurate diagnosis, it is important to query the KG to provide interactive explanations about learned biomarkers. Inspired by these, we construct a domain-specific KG, particularly for cancer-specific biomarker discovery. The KG is constructed by integrating cancer-related knowledge and facts from multiple sources. First, we construct a domain-specific ontology, which we call OncoNet Ontology (ONO). The ONO ontology is developed to enable semantic reasoning for verification of the predictions for relations between diseases and genes. The KG is then developed and enriched by harmonizing the ONO, additional metadata schemas, ontologies, controlled vocabularies, and additional concepts from external sources using a BERT-based information extraction method. BioBERT and SciBERT are finetuned with the selected articles crawled from PubMed. We listed down some queries and some examples of QA and deducing knowledge based on the KG.

Md Shajalal, Sebastian Denef, Md. Rezaul Karim et al.

Recent technological advancements have led to a large number of patents in a diverse range of domains, making it challenging for human experts to analyze and manage. State-of-the-art methods for multi-label patent classification rely on deep neural networks (DNNs), which are complex and often considered black-boxes due to their opaque decision-making processes. In this paper, we propose a novel deep explainable patent classification framework by introducing layer-wise relevance propagation (LRP) to provide human-understandable explanations for predictions. We train several DNN models, including Bi-LSTM, CNN, and CNN-BiLSTM, and propagate the predictions backward from the output layer up to the input layer of the model to identify the relevance of words for individual predictions. Considering the relevance score, we then generate explanations by visualizing relevant words for the predicted patent class. Experimental results on two datasets comprising two-million patent texts demonstrate high performance in terms of various evaluation measures. The explanations generated for each prediction highlight important relevant words that align with the predicted class, making the prediction more understandable. Explainable systems have the potential to facilitate the adoption of complex AI-enabled methods for patent classification in real-world applications.

Md. Rezaul Karim, Felix Hermsen, Sisay Adugna Chala et al.

Money laundering is the process where criminals use financial services to move massive amounts of illegal money to untraceable destinations and integrate them into legitimate financial systems. It is very crucial to identify such activities accurately and reliably in order to enforce an anti-money laundering (AML). Despite tremendous efforts to AML only a tiny fraction of illicit activities are prevented. From a given graph of money transfers between accounts of a bank, existing approaches attempted to detect money laundering. In particular, some approaches employ structural and behavioural dynamics of dense subgraph detection thereby not taking into consideration that money laundering involves high-volume flows of funds through chains of bank accounts. Some approaches model the transactions in the form of multipartite graphs to detect the complete flow of money from source to destination. However, existing approaches yield lower detection accuracy, making them less reliable. In this paper, we employ semi-supervised graph learning techniques on graphs of financial transactions in order to identify nodes involved in potential money laundering. Experimental results suggest that our approach can sport money laundering from real and synthetic transaction graphs.

Md Muhtasim Munif Fahim, Soyda Humyra Yesmin, Saiful Islam et al.

We introduce Green-NAS, a multi-objective NAS (neural architecture search) framework designed for low-resource environments using weather forecasting as a case study. By adhering to 'Green AI' principles, the framework explicitly minimizes computational energy costs and carbon footprints, prioritizing sustainable deployment over raw computational scale. The Green-NAS architecture search method is optimized for both model accuracy and efficiency to find lightweight models with high accuracy and very few model parameters; this is accomplished through an optimization process that simultaneously optimizes multiple objectives. Our best-performing model, Green-NAS-A, achieved an RMSE of 0.0988 (i.e., within 1.4% of our manually tuned baseline) using only 153k model parameters, which is 239 times fewer than other globally applied weather forecasting models, such as GraphCast. In addition, we also describe how the use of transfer learning will improve the weather forecasting accuracy by approximately 5.2%, in comparison to a naive approach of training a new model for each city, when there is limited historical weather data available for that city.

Md. Rezaul Karim, Sumon Kanti Dey, Tanhim Islam et al.

The exponential growths of social media and micro-blogging sites not only provide platforms for empowering freedom of expressions and individual voices, but also enables people to express anti-social behaviour like online harassment, cyberbullying, and hate speech. Numerous works have been proposed to utilize textual data for social and anti-social behaviour analysis, by predicting the contexts mostly for highly-resourced languages like English. However, some languages are under-resourced, e.g., South Asian languages like Bengali, that lack computational resources for accurate natural language processing (NLP). In this paper, we propose an explainable approach for hate speech detection from the under-resourced Bengali language, which we called DeepHateExplainer. Bengali texts are first comprehensively preprocessed, before classifying them into political, personal, geopolitical, and religious hates using a neural ensemble method of transformer-based neural architectures (i.e., monolingual Bangla BERT-base, multilingual BERT-cased/uncased, and XLM-RoBERTa). Important(most and least) terms are then identified using sensitivity analysis and layer-wise relevance propagation(LRP), before providing human-interpretable explanations. Finally, we compute comprehensiveness and sufficiency scores to measure the quality of explanations w.r.t faithfulness. Evaluations against machine learning~(linear and tree-based models) and neural networks (i.e., CNN, Bi-LSTM, and Conv-LSTM with word embeddings) baselines yield F1-scores of 78%, 91%, 89%, and 84%, for political, personal, geopolitical, and religious hates, respectively, outperforming both ML and DNN baselines.

Md. Rezaul Karim, Bharathi Raja Chakravarthi, John P. McCrae et al.

Exponential growths of social media and micro-blogging sites not only provide platforms for empowering freedom of expressions and individual voices but also enables people to express anti-social behaviour like online harassment, cyberbullying, and hate speech. Numerous works have been proposed to utilize these data for social and anti-social behaviours analysis, document characterization, and sentiment analysis by predicting the contexts mostly for highly resourced languages such as English. However, there are languages that are under-resources, e.g., South Asian languages like Bengali, Tamil, Assamese, Telugu that lack of computational resources for the NLP tasks. In this paper, we provide several classification benchmarks for Bengali, an under-resourced language. We prepared three datasets of expressing hate, commonly used topics, and opinions for hate speech detection, document classification, and sentiment analysis, respectively. We built the largest Bengali word embedding models to date based on 250 million articles, which we call BengFastText. We perform three different experiments, covering document classification, sentiment analysis, and hate speech detection. We incorporate word embeddings into a Multichannel Convolutional-LSTM (MConv-LSTM) network for predicting different types of hate speech, document classification, and sentiment analysis. Experiments demonstrate that BengFastText can capture the semantics of words from respective contexts correctly. Evaluations against several baseline embedding models, e.g., Word2Vec and GloVe yield up to 92.30%, 82.25%, and 90.45% F1-scores in case of document classification, sentiment analysis, and hate speech detection, respectively during 5-fold cross-validation tests.

Md. Rezaul Karim, Till Döhmen, Dietrich Rebholz-Schuhmann et al.

Amid the coronavirus disease(COVID-19) pandemic, humanity experiences a rapid increase in infection numbers across the world. Challenge hospitals are faced with, in the fight against the virus, is the effective screening of incoming patients. One methodology is the assessment of chest radiography(CXR) images, which usually requires expert radiologist's knowledge. In this paper, we propose an explainable deep neural networks(DNN)-based method for automatic detection of COVID-19 symptoms from CXR images, which we call DeepCOVIDExplainer. We used 15,959 CXR images of 15,854 patients, covering normal, pneumonia, and COVID-19 cases. CXR images are first comprehensively preprocessed, before being augmented and classified with a neural ensemble method, followed by highlighting class-discriminating regions using gradient-guided class activation maps(Grad-CAM++) and layer-wise relevance propagation(LRP). Further, we provide human-interpretable explanations of the predictions. Evaluation results based on hold-out data show that our approach can identify COVID-19 confidently with a positive predictive value(PPV) of 91.6%, 92.45%, and 96.12%; precision, recall, and F1 score of 94.6%, 94.3%, and 94.6%, respectively for normal, pneumonia, and COVID-19 cases, respectively, making it comparable or improved results over recent approaches. We hope that our findings will be a useful contribution to the fight against COVID-19 and, in more general, towards an increasing acceptance and adoption of AI-assisted applications in the clinical practice.

Md. Rezaul Karim, Michael Cochez, Oya Beyan et al.

The discovery of important biomarkers is a significant step towards understanding the molecular mechanisms of carcinogenesis; enabling accurate diagnosis for, and prognosis of, a certain cancer type. Before recommending any diagnosis, genomics data such as gene expressions(GE) and clinical outcomes need to be analyzed. However, complex nature, high dimensionality, and heterogeneity in genomics data make the overall analysis challenging. Convolutional neural networks(CNN) have shown tremendous success in solving such problems. However, neural network models are perceived mostly as `black box' methods because of their not well-understood internal functioning. However, interpretability is important to provide insights on why a given cancer case has a certain type. Besides, finding the most important biomarkers can help in recommending more accurate treatments and drug repositioning. In this paper, we propose a new approach called OncoNetExplainer to make explainable predictions of cancer types based on GE data. We used genomics data about 9,074 cancer patients covering 33 different cancer types from the Pan-Cancer Atlas on which we trained CNN and VGG16 networks using guided-gradient class activation maps++(GradCAM++). Further, we generate class-specific heat maps to identify significant biomarkers and computed feature importance in terms of mean absolute impact to rank top genes across all the cancer types. Quantitative and qualitative analyses show that both models exhibit high confidence at predicting the cancer types correctly giving an average precision of 96.25%. To provide comparisons with the baselines, we identified top genes, and cancer-specific driver genes using gradient boosted trees and SHapley Additive exPlanations(SHAP). Finally, our findings were validated with the annotations provided by the TumorPortal.

Md. Rezaul Karim, Michael Cochez, Joao Bosco Jares et al.

Interference between pharmacological substances can cause serious medical injuries. Correctly predicting so-called drug-drug interactions (DDI) does not only reduce these cases but can also result in a reduction of drug development cost. Presently, most drug-related knowledge is the result of clinical evaluations and post-marketing surveillance; resulting in a limited amount of information. Existing data-driven prediction approaches for DDIs typically rely on a single source of information, while using information from multiple sources would help improve predictions. Machine learning (ML) techniques are used, but the techniques are often unable to deal with skewness in the data. Hence, we propose a new ML approach for predicting DDIs based on multiple data sources. For this task, we use 12,000 drug features from DrugBank, PharmGKB, and KEGG drugs, which are integrated using Knowledge Graphs (KGs). To train our prediction model, we first embed the nodes in the graph using various embedding approaches. We found that the best performing combination was a ComplEx embedding method creating using PyTorch-BigGraph (PBG) with a Convolutional-LSTM network and classic machine learning-based prediction models. The model averaging ensemble method of three best classifiers yields up to 0.94, 0.92, 0.80 for AUPR, F1-score, and MCC, respectively during 5-fold cross-validation tests.

Md. Rezaul Karim, Michael Cochez, Achille Zappa et al.

The study of genetic variants can help find correlating population groups to identify cohorts that are predisposed to common diseases and explain differences in disease susceptibility and how patients react to drugs. Machine learning algorithms are increasingly being applied to identify interacting GVs to understand their complex phenotypic traits. Since the performance of a learning algorithm not only depends on the size and nature of the data but also on the quality of underlying representation, deep neural networks can learn non-linear mappings that allow transforming GVs data into more clustering and classification friendly representations than manual feature selection. In this paper, we proposed convolutional embedded networks in which we combine two DNN architectures called convolutional embedded clustering and convolutional autoencoder classifier for clustering individuals and predicting geographic ethnicity based on GVs, respectively. We employed CAE-based representation learning on 95 million GVs from the 1000 genomes and Simons genome diversity projects. Quantitative and qualitative analyses with a focus on accuracy and scalability show that our approach outperforms state-of-the-art approaches such as VariantSpark and ADMIXTURE. In particular, CEC can cluster targeted population groups in 22 hours with an adjusted rand index of 0.915, the normalized mutual information of 0.92, and the clustering accuracy of 89%. Contrarily, the CAE classifier can predict the geographic ethnicity of unknown samples with an F1 and Mathews correlation coefficient(MCC) score of 0.9004 and 0.8245, respectively. To provide interpretations of the predictions, we identify significant biomarkers using gradient boosted trees(GBT) and SHAP. Overall, our approach is transparent and faster than the baseline methods, and scalable for 5% to 100% of the full human genome.