https://hdl.handle.net/20.500.12573/395 Scopus Indexed Publications Collection 2026-05-08T05:19:05Z https://hdl.handle.net/20.500.12573/2542 An effective colorectal polyp classification for histopathological images based on supervised contrastive learning Yengec-Tasdemir,Sena Busra; Aydin,Zafer; Akay,Ebru; Doğan,Serkan; Yilmaz,Bulent Early detection of colon adenomatous polyps is pivotal in reducing colon cancer risk. In this context, accurately distinguishing between adenomatous polyp subtypes, especially tubular and tubulovillous, from hyperplastic variants is crucial. This study introduces a cutting-edge computer-aided diagnosis system optimized for this task. Our system employs advanced Supervised Contrastive learning to ensure precise classification of colon histopathology images. Significantly, we have integrated the Big Transfer model, which has gained prominence for its exemplary adaptability to visual tasks in medical imaging. Our novel approach discerns between in-class and out-of-class images, thereby elevating its discriminatory power for polyp subtypes. We validated our system using two datasets: a specially curated one and the publicly accessible UniToPatho dataset. The results reveal that our model markedly surpasses traditional deep convolutional neural networks, registering classification accuracies of 87.1% and 70.3% for the custom and UniToPatho datasets, respectively. Such results emphasize the transformative potential of our model in polyp classification endeavors 2024-01-01T00:00:00Z https://hdl.handle.net/20.500.12573/2541 Matching variants for functional characterization of genetic variants Cevik,Sabiha; Zhao,Pei; Zorluer,Atiyye; Pir, Mustafa S.; Bian, Wenyin; Kaplan, Oktay I. Rapid and low-cost sequencing, as well as computer analysis, have facilitated the diagnosis of many genetic diseases, resulting in a substantial rise in the number of disease-associated genes. However, genetic diagnosis of many disorders remains problematic due to the lack of interpretation for many genetic variants, especially missenses, the infeasibility of high-throughput experiments on mammals, and the shortcomings of computational prediction technologies. Additionally, the available mutant databases are not well-utilized. Toward this end, we used Caenorhabditis elegans mutant resources to delineate the functions of eight missense variants (V444I, V517D, E610K, L732F, E817K, H873P, R1105K, and G1205E) and two stop codons (W937stop and Q1434stop), including several matching variants (MatchVar) with human in ciliopathy associated IFT-140 (also called CHE-11)//IFT140 (intraflagellar transport protein 140). Moreover, MatchVars carrying C. elegans mutants, including IFT-140(G680S) and IFT-140(P702A) for the human (G704S) (dbSNP: rs150745099) and P726A (dbSNP: rs1057518064 and a conflicting variation) were created using CRISPR/Cas9. IFT140 is a key component of IFT complex A (IFT-A), which is involved in the retrograde transport of IFT along cilia and the entrance of G protein-coupled receptors into cilia. Functional analysis of all 10 variants revealed that P702A and W937stop, but not others phenocopied the ciliary phenotypes (short cilia, IFT accumulations, mislocalization of membrane proteins, and cilia entry of nonciliary proteins) of the IFT-140 null mutant, indicating that both P702A and W937stop are phenotypic in C. elegans. Our functional data offered experimental support for interpreting human variants, by using ready-to-use mutants carrying MatchVars and generating MatchVars with CRISPR/Cas9. 2023-01-01T00:00:00Z https://hdl.handle.net/20.500.12573/2540 Integrating Biological Domain Knowledge with Machine Learning for Identifying Colorectal-Cancer-Associated Microbial Enzymes in Metagenomic Data Bakir-Gungor, Burcu; Ersoz, Nur Sebnem; Yousef, Malik Advances in metagenomics have revolutionized our ability to elucidate links between the microbiome and human diseases. Colorectal cancer (CRC), a leading cause of cancer-related mortality worldwide, has been associated with dysbiosis of the gut microbiome. This study aims to develop a method for identifying CRC-associated microbial enzymes by incorporating biological domain knowledge into the feature selection process. Conventional feature selection techniques often evaluate features individually and fail to leverage biological knowledge during metagenomic data analysis. To address this gap, we propose the enzyme commission (EC)-nomenclature-based Grouping-Scoring-Modeling (G-S-M) method, which integrates biological domain knowledge into feature grouping and selection. The proposed method was tested on a CRC-associated metagenomic dataset collected from eight different countries. Community-level relative abundance values of enzymes were considered as features and grouped based on their EC categories to provide biologically informed groupings. Our findings in randomized 10-fold cross-validation experiments imply that glycosidases, CoA-transferases, hydro-lyases, oligo-1,6-glucosidase, crotonobetainyl-CoA hydratase, and citrate CoA-transferase enzymes can be associated with CRC development as part of different molecular pathways. These enzymes are mostly synthesized by Eschericia coli, Salmonella enterica, Klebsiella pneumoniae, Staphylococcus aureus, Streptococcus pneumoniae, and Clostridioides dificile. Comparative evaluation experiments showed that the proposed model consistently outperforms traditional feature selection methods paired with various classifiers. 2025-01-01T00:00:00Z https://hdl.handle.net/20.500.12573/2539 Axial free vibration analysis of a tapered nanorod using Adomian decomposition method Coskun, Safa B.; Kara, Ozge; Atay, Mehmet T. This study aimed to conduct an analysis of the axial free vibration of tapered nanorods based on nonlocal elasticity theory. The small-scale effect on the free axial vibration of a tapered nanorod was studied employing the Adomian decomposition method (ADM) and the finite difference method (FDM) as a checking tool where a contradiction existed between the results of this study and available results in one highly cited work in the literature, which was used for comparison purposes in this work. Different boundary conditions for the nanorod were considered: fixed-fixed nanorod, fixed-free nanorod, and fixed-linear spring nanorod. The governing equation of the problem is a variable coefficient differential equation for which analytical solutions are strictly limited. For this type of problem, analytical approximate methods are effective, and there are many studies available in the literature on the application of these methods to solve linear/nonlinear ordinary/partial differential equations. ADM is one of the methods and was successfully used in this study to analyze the free vibration of nanorods. The results obtained in this study have shown that the presented technique is so powerful and has potential for applications in nanomechanics based on nonlocal elasticity theory. 2025-01-01T00:00:00Z