4
Archived Theses
0
DOIs Assigned
0%
DOI Rate
Discipline
The investigation of mechanistic differences of Rac1P29S and Rac1A159V activation via molecular dynamics simulations
Rac1 is a small GTPase which plays key roles in actin reorganization, cell motility, cell survival/growth as well as in various cancer types and neurodegenerative diseases. Similar to other Ras superfamily GTPases, Rac1 switches between GTP-bound active and GDP-bound inactive states. When active, Rac1 signals to various downstream effectors including Pak family kinases. The switch I and switch II regions open and close during GDP/GTP exchange. Rac1P29S and Rac1A159V (paralogous to K-RasA146) mutations are the two most common somatic mutations of Rac1. Rac1P29S is a known hotspot for melanoma, where it is the third most occurring mutation after B-RafV600 and N-RasQ61 mutations. Rac1A159V is most commonly observed in head and neck cancer. Both mutations are relatively newly discovered and require better characterization. In this thesis, how the mutations Rac1P29S and Rac1A159V differ the Rac1 dynamics is investigated by using molecular dynamics simulations. A total of five systems are simulated as follows: Rac1WT-GTP, Rac1WT-GDP, Rac1P29S-GTP, Rac1P29S-GDP, and Rac1A159V-GTP. Here, wild-type systems are considered as the control groups. For the analysis of the simulation trajectories, we focused on the conformational changes of switch regions and changes in nucleotide binding residues as these changes are important for GDP/GTP exchange of Rac1. This thesis suggests that P29S and A159V mutations activate Rac1 with different mechanisms. In the Rac1P29S-GTP system, proline to serine substitution changes the flexibility of the switch I region and keeps the switch in an open conformation. We propose that the open conformation of switch I region is one of the underlying reasons for rapid GDP/GTP exchange of Rac1P29S. On the other hand, in Rac1A159V-GTP, some of the contacts of guanosine ring of GTP with Rac1 temporarily lost, enabling guanosine ring to move towards switch I region and subsequently close the switch. The switch II regions of both Rac1P29S-GTP and Rac1A159V-GTP systems are stabilized in a closed conformation with respect to the Rac1WT-GTP system. Rac1A159V-GTP adopts a conformation similar to Ras state 2, where both switch regions are in closed conformations, switch I residue Thr35 forms a hydrogen bond with the nucleotide, and switch II residue Gly60 also interacts with the nucleotide. The fact that all Rac1WT-GTP, Rac1P29S-GTP, and Rac1A159V-GTP are stabilized with different conformations suggests that all three systems would interact with Rac1 regulators and downstream effectors differently.
Makine öğrenmesi bazlı sınıflandırma kullanarak nörogelişimsel hastalıklar için ayırt edici özelliklerin bulunması
Major Depressive Disorder (MDD) is one of the leading neurodevelopmental disorders worldwide and its heterogeneous and complex nature remains a significant challenge for scientists to fully understand and unravel. In recent years, the importance of gut microbiome through the human gut-brain axis has gathered the attention of scientists to analyze and model the bacterial components of the human gut and its effects on the human brain, especially in the case of MDD. In this study, we analyzed the American Gut Project (AGP) dataset using fecal samples of 361 controls and 23 MDD patients. After retrieving the Qiita bioinformatics analysis, the cohort was analyzed for its characteristics and for alpha and beta diversity indexes which did not reveal any statistical significance except for the age of patients. Various differential abundance analysis (DAA) methods were conducted to find potential biomarkers and these results were fed into our machine learning models as an alternative DAA-filtered dataset compared to the raw dataset to find important features. Our best two models, which were Random Forest and XGBoost, at their intersection, have found that despite some inconsistencies in the literature, species Bifidobacterium adolescentis and genera Odoribacter, Ruminococcus, and Adlercreutzia could be potential biomarkers for MFD. Our models found that species B. adolescentis decreased in MDD patients whereas the rest increased in MDD patients. Despite supporting evidence for these potential biomarkers, factors such as modeling and filtering choices, as well as external influences like sex, stress and diet should also be taken into account when analyzing gut microbiome datasets.
Bağlayıcı püskürtme tekniğinde adaptif dilimleme yöntemi ve uygulama teknolojisi
Eklemeli imalatın, klasik yöntemlerle üretilmesi mümkün olmayan karmaşık parçaları üretebilmesi, daha az malzeme kullanılması, yerinde üretim olanağıyla tedarik zincirini rahatlatması, her türlü malzeme ile üretim yapabilmesi ve daha hafif parça üretebilme gibi birçok avantajından dolayı çoğu endüstri kolunda kullanımı ciddi anlamda artmaktadır. Eklemeli imalat yöntemlerinden olan ve bu tez kapsamında araştırılan bağlayıcı püskürtme tekniğinin ise birçok ekonomik rapora göre önümüzdeki on yıl içerisinde en çok büyüyecek eklemeli imalat yöntemi olacağı öngörülmektedir. Eklemeli imalat yöntemlerinin birçok avantajı bulunmasına rağmen üretim hızı bakımından klasik imalat yöntemlerine göre yavaş kalabilmektedir. Bunun için bu çalışmada bağlayıcı püskürtme yönteminde imalat hızının arttırılması amaçlanmıştır. İmalat hızının arttırılması için bağlayıcı püskürtmede adaptif dilimlemenin kullanılması önerilmiştir. Adaptif dilimlemenin bağlayıcı püskürtme tekniğinde verimli bir şekilde kullanılabilmesi için değişken bağlayıcı miktarı algoritması (DBMA) geliştirilmiştir. Yapılan ön çalışmalarda DBMA, tecrübeye ve literatüre dayanan parametre verileri ile denenmiş olup başarılı sonuçlar elde edilmiştir. Ancak DBMA'da özellikle katman kalınlığının ve doygunluk oranın optimize edilmesi gerekmektedir. Bunun için Taguchi yöntemi kullanılarak, adaptif dilimlemede katman kalınlığı ve doygunluk oranı optimizasyonu yapılmıştır. Taguchi deney tasarımına göre 9 farklı koşulda, 3'er tekerrür olmak üzere toplam 27 adet numune üretilmiştir. Numunelerin ham hallerinin en boyutunun ölçüsüne bakılmıştır. Sonrasında numuneler 1500°C'de 2 saat sinterlenmiştir. Sinterleme sonrası yüzey pürüzlülük ve yoğunluk testleri yapılmıştır. Bu testlerin sonucunda optimum baskı koşulu olarak katman kalınlığı için 180-250 µm, doygunluk için ise %50 değerine karar verilmiştir. Sonrasında adaptif dilimlemeyi uygulamak için ayrı bir test numunesi tasarlanmıştır. Bu test numunesi belirlenen parametreler ile adaptif (180 µm-250 µm), ince katman (180 µm) ve kalın katman (250 µm) olmak üzere 3 adet üretilmiştir. Adaptif dilimlenmiş numune ile ince katmanlı numunenin pürüzlülük değerleri birbiri ile benzer olup kalın katmanlı numuneden daha iyi çıkmıştır. Adaptif numunede ince katmanlı numuneye göre %12,31 daha az katman kullanılarak benzer sonuç elde edilmiştir. Dolayısıyla kullanılan yöntemlerin başarılı olduğu ıspatlanmıştır. Ayrıca yüzey pürüzlülük ölçümlerini daha kolay ve ekonomik yapmak adına görüntü işleme ile ölçüm olanakları araştırılmıştır. Görüntü işleme verileri ile pürüzlülük ölçüm cihazıyla alınan veriler birbirleri ile uyumlu çıkmıştır.
Ses ve görüntü dönüşümü kullanilarak android kötücül yazilim tespiti
Mobile devices have started a new era with their hardware and various software developed for them. A security vulnerability in these devices could lead to the theft of personal information, breaches of privacy, and even financial loss. Therefore, ensuring that the apps downloaded to the devices are reliable and safe is very important. For this purpose, within the scope of this thesis, an investigation into the image and audio-based approaches for Android malware detection and family classification is conducted. In the image-based approach, an end-to-end method is proposed that treats Android application files as binary sequences. In the method, grayscale image representations were created for each sample, and training and testing processes were carried out with CNN. Image representations of malware can be made by treating the files as binary sequences or the extracted static features in matrix form. For this reason, the impact of static feature set combinations on classification performance is also investigated. Initially, all possible combinations of four different feature sets obtained from Android application files are considered, and their effects on classification performance are investigated. Effective feature set combinations are determined by evaluating all combinations with ten different classification algorithms. Subsequently, RGB images are created with feature set combinations, and training and testing processes are carried out using CNN. In the results obtained, it was seen that with different feature set combinations, a performance above 99% could be obtained. In addition to image representations of Android applications, audio representations can also be created. Although audio-based approaches are less common than image-based ones in the literature, they can achieve similarly high classification accuracies. In this context, an audio-based method is proposed, treating the Android malware family detection problem as a music category classification problem. Android application files were converted to audio files, and their audio-based attributes were extracted. Then, features with high discrimination were determined with four different feature selection algorithms, and the classification processes were carried out. Family detection was performed with 96.6% accuracy in experiments on an eight-class data set. At the end of the thesis, discussions were made about the methods used in the study.