Browsing by Author "Hosen, Md. Faruk"

Now showing 1 - 7 of 7

An LSTM network-based model with attention techniques for predicting linear T-cell epitopes of the hepatitis C virus
(Scopus, 2024) Hosen, Md. Faruk; Mahmud, S. M. Hasan; Goh, Kah Ong Michael; Uddin, Muhammad Shahin; Nandi, Dip; Shatabda, Swakkhar; Shoombuatong, Watshara
: Hepatitis C virus (HCV) infection remains a significant global health challenge, often resulting in severe longterm physical complexity and even death. Since its discovery, HCV has exhibited substantial genetic variability, complicating vaccine development. Although some therapeutic approach have shown efficacy against certain HCV genotypes, a universally effective vaccine is still lacking. Recent research suggests that the body’s cellular immune response, particularly T cell epitopes of HCV (TCE-HCVs), plays a vital role in fighting the virus. Therefore, the precise and rapid identification of TCE-HCVs is essential for chronic HCV infection. In this work, we proposed a novel TCE-HCVs prediction model AttLSTM, which combines attention mechanism and long shortterm memory (LSTM). Specifically, we employed four robust feature encoding techniques: One-Hot Encoding, Global Vectors (GloVe), fastText, and Word2Vec to encode protein sequences. Additionally, k-mer embedding was utilized to help the model identify significant subsequence fragments within the protein sequences. To optimize the model’s performance, irrelevant features are eliminated using the SHapley Additive exPlanations (SHAP) approach. The resulting optimal feature subset was then fed into the AttLSTM model to identify TCEHCVs. The attention mechanism in this model dynamically captures the pairwise correlations of each neighboring target pair within a sliding window, thereby enhancing the understanding of the local environment of target residues. Extensive experiments showed that AttLSTM outperformed conventional machine learning (ML) classifiers in predictive performance. Notably, in k-fold cross validation, AttLSTM achieved superior performance compared to existing methods with accuracy of 80.77 %, MCC of 0.632, and AUC of 0.891. This exceptional performance indicates that AttLSTM has a strong predictive capability for identifying TCE-HCVs. We anticipate that AttLSTM will expedite the rapid identification of promising TCE-HCVs, aiding in the development of diagnostic and immunotherapeutic treatments for HCV in the future.
Identification of Drug and Protein-Protein Interaction Network Among Stress and Depression: A Bioinformatics Approach
(Elsevier, 2023-01-18) Basar, Md. Abul; Hosen, Md. Faruk; Paul, Bikash Kumar; Hasan, Md. Rakibul; Shamim, S.M.; Bhuyian, Touhid
The fields of data mining, computational biology, and statistics have been combined to form the massive research area of bioinformatics. In the areas of genetics, education, and healthcare, bioinformatics integrates the tools available in different fields such as computing, inventorying, performing statistical analyses, and collecting and processing genomic data. Stress and depression are two of the most severe mental disorders that affect people of all ages, including children and adults. The goal of this study was to look into the relationship between genetic alterations and the two diseases mentioned above as well as to develop a PPI network or related channel. The first step is to determine whether or not there is a biological relationship between them. This would assist us in connecting both of them as well as building therapeutic drugs that are effective against stress and depression disorders. Using R programs, the genes that are responsible for different diseases are acquired, pre-processed, analyzed, and mined in order to better understand them. During the study, a novel pathway was discovered. Based on common genes between the two diseases studied, the PPI network, gene-miRNA interaction, TF-gene interaction, and PDI network were established. This data can help us better understand how the PPI network binds to its ligands. We anticipate that our study will contribute to the development of new drugs for stress and depression.
Identification of Key Signaling Pathways and Novel Computational Drug Target for Depression and Coronary Artery Disease
(2024-12-19) Hosen, Md. Faruk; Abul Basar, Md.; Shahin Uddin, Muhammad; Yasmin, Mst. Farjana; Morshed, Monir
Psychological disorders, such as anxiety, bipolar disorder, panic disorder, stress, depression, and schizophrenia, are increasingly prevalent worldwide. Among these conditions, depression is particularly notable as one of the most common and debilitating neuropsychiatric disorders. Individuals with depression may be at a higher risk of developing oronary Artery Disease (CAD). Depression can contribute to poor lifestyle choices, such as unhealthy eating, lack of exercise, and smoking, which are risk factors for CAD. The emotional stress and anxiety associated with depression can strain the heart and exacerbate CAD symptoms. The relationship is not one-sided. CAD itself can be a significant source of emotional distress, leading to symptoms of depression and anxiety in affected individuals. Managing both conditions in tandem can be complicated. Treating CAD may involve medications, lifestyle modifications, and potentially surgical interventions. Meanwhile, depression often requires therapy, counseling, and medication. Coordinating care and addressing both conditions simultaneously is crucial. In our study, we investigated the molecular connections between CAD and Depression using GSE98793 and GSE20681 microarray datasets. After preprocessing, we identified key hub genes, including CCT2, SVIL, REPS2, ASPH, and UBC, in the shared ProteinProtein Interaction network. KEGG pathways linked these DEGs to colorectal and cancer pathways. Our next steps involve exploring microRNAs, TFs, and GO analysis. These findings offer promising leads for potential therapies, uniting CAD and Depression under a common molecular framework, advancing our understanding of these conditions.
Identification of key signaling pathways and novel computational drug target for oral cancer, metabolic disorders and periodontal disease
(2024-12-15) Alam, Mohammad Khursheed; Hosen, Md. Faruk; Ganji, Kiran Kumar; Ahmed, Kawsar; M. Bui, Francis
Due to conventional endocrinological methods, there is presently no shared work available, and no therapeutic options have been demonstrated in oral cancer (OC) and periodontal disease (PD), type 2 diabetes (T2D), and obese patients. The aim of this study is to determine the similar molecular pathways and potential therapeutic targets in PD, OC, T2D, and obesity that may be used to anticipate the progression of the disease. Four Gene Expression Omnibus (GEO) microarray datasets (GSE29221, GSE15773, GSE16134, and GSE13601) are used for finding differentially expressed genes (DEGs) for T2D, obese, and PD patients with OC in order to explore comparable pathways and therapeutic medications. Gene ontology (GO) and pathway analysis were used to investigate the functional annotations of the genes. The hub genes were then identified using protein-protein interaction (PPI) networks, and the most significant PPI components were evaluated using a clustering approach. These three gene expression-based datasets yielded a total of seven common DEGs. According to the GO annotation, the majority of the DEGs were connected with the microtubule cytoskeleton structure involved in mitosis. The KEGG pathways revealed that the concordant DEGs are connected to the cell cycle and progesterone-mediated oocyte maturation. Based on topological analysis of the PPI network, major hub genes (CCNB1, BUB1, TTK, PLAT, and AHNAK) and notable modules were revealed. This work additionally identified the connection of TF genes and miRNAs with common DEGs, as well as TF activity. Predictive drug analysis yielded concordant drug compounds involved with T2D, OC, PD, and obesity disorder, which might be beneficial for examining the diagnosis, treatment, and prognosis of metabolic disorders and Oral cancer.
Integrated Bioinformatics and Machine Learning Analysis Reveals Shared Key Candidate Biomarkers and Therapeutic Targets in Ulcerative Colitis and Colorectal Cancer
(2024-10-24) Sarker, Sakib; Hosen, Md. Faruk; Abul Basar, Md.; Ahammed, Emon
The interplay between ulcerative colitis (UC) and colorectal cancer (CRC) has garnered significant research interest due to their potential shared molecular mechanisms. This study aims to identify common significant biomarkers and potential therapeutic targets for UC and CRC. We utilized two microarray datasets to perform differential expression analysis, identifying DEGs for both conditions. Subsequent ML-based gene selection was conducted using SHapley Additive exPlanations (SHAP) algorithm models on the respective datasets. Common ML-based DEGs were then identified and a protein-protein interaction (PPI) network was constructed using the STRING database. The PPI network was visualized and analyzed in Cytoscape, with the top ten hub genes identified using the Degree method in the cytoHubba plugin. The hub genes identified were CDC20, ANLN, HMMR, CCNB1, CDK1, KIF20A, ECT2, KIF11, NUF2, and CCNA2. These genes were further validated through survival analysis, establishing their significance in patient outcomes. Finally, we explored the drug-gene interaction network to identify potential therapeutic drugs targeting these hub genes. This comprehensive bioinformatics approach provides insights into the shared molecular pathways in UC and CRC and highlights poten- tial therapeutic targets for future research and drug development.
Integrated Bioinformatics and Machine Learning Analysis Uncovers Key Pathways and Therapeutic Targets for Hypertension and Chronic Kidney Disease
(2024-12-20) Wasima, Jeba; Hosen, Md. Faruk; D Cruze, Francis Rudra; Shahin Uddin, Muhammad
Hypertension is a serious cardiovascular disease that substantially raises morbidity and mortality rates worldwide. People who have high blood pressure have been found to have an increased risk of developing chronic kidney disease (CKD) in recent years. The goal of this research is to use modern bioinformatics approaches to find potential treatment candidates and clarify the underlying biological pathways linked to both hypertension and CKD. Sample from individuals with CKD and hypertension were taken from two publicly available microarray datasets, GSE33463 and GSE66494. Consistent differentially expressed genes (DEGs) were found following thorough pre- processing and Python analysis. A Venn diagram was used to show where these DEGs’ regulatory crossings were. The most functionally important genes were then identified via topological analysis after protein-protein interaction (PPI) networks were built. UBC, ARRIB1, FADD and EIF3D have been identified as important hub genes. These concordant DEGs are tightly linked to the Toll-like receptor signaling pathway, which is a crucial mechanism in the control of the immunological response, according to pathway enrichment analysis performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG).In order to better understand gene relationships, future research will examine modular network studies, transcription factor (TF), microRNA (miRNA) network regulation, and gene ontology (GO) analysis. Concordant DEGs have been used to select a number of possible medicinal molecules, providing a promising path forward for therapeutic research.
MN1 DN121 Vg19: A Multi-stream Ensemble Model for Detecting External Damage on Tomato Surface
(2025-06-12) Juthy, Most. Jebun Nahar; Tauhid; Hosen, Md. Faruk; Abul Basara, Md.; Aktar, Mst. Nargis; Shahin Uddin, Muhammad
The use of advanced computer vision and powerful deep-learning capabilities is increasingly important in the agricultural industry, particularly for monitoring crops and detecting diseases. However, detecting external damage to fruit crops remains a significant challenge. In our study, we introduced a multi-stream ensemble model called MN1_DN121_Vg19, which is designed to identify external damage on tomato surfaces. We separately used different pre-trained approaches, including InceptionV3, ResNet50, ResNet101, VGG16, VGG19M obileNetV1, and DenseNet121, and we examined how well they performed. After evaluating their performance, We merged the three wellknown pre-trained methods: MobileNetV1, DenseNet121, and VGG19, and extracted the learned features from these models. After that, we fine-tuned some of the top layers to effectively learn the features from our used dataset. The top 10 layers of DenseNet121 and MobileNet, as well as the top 4 layers of VGG19, were fine-tuned to create the merged model. The accuracy of the suggested multi-stream ensemble model was 98.51% using a dataset of 6500 images across 4 classes. When compared to existing pre-trained models, our merged model demonstrated superior performance across all metrics.