Yüksek Lisans Tezleri
Permanent URI for this collectionhttps://hdl.handle.net/20.500.12573/5799
Browse
14 results
Search Results
Master Thesis Tree-net: Biyomedikal Görüntü Segmentasyonu için Tree-net: Darboğaz Özellik Süpervizyonu Kullanılan Yapay Sinir Ağı Modeli(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2024) Demirci, Orhan; Yılmaz, BülentIn this thesis, we introduce Tree-NET, a novel approach for medical image segmentation utilizing bottleneck feature supervision. This method enhances traditional segmentation algorithms by keeping supervision between bottleneck features of the network. The primary goal is to improve the model's ability to learn discriminative and robust features while simultaneously reducing computational costs. Bottleneck feature supervision involves compressing the input and label data using Autoencoders and then supervising the bottleneck features with a segmentation network named 'Bridge-Net,' which can be any segmentation model of choice. We applied Tree-NET to two critical medical image segmentation tasks: skin lesion segmentation and polyp segmentation. Our experiments demonstrate significant improvements in segmentation accuracy and efficiency. For instance, the U-NET backboned Tree-NET uses only 154.43 MB for executing and storing the model, which is almost 3.5 times smaller than the original U-Net while having a close number of trainable parameters. In skin lesion segmentation, Tree-NET achieved dice, Intersection-over-Union (IoU), and accuracy scores of 0.893, 0.751, and 0.977 respectively. For polyp segmentation, the scores were 0.856, 0.795, and 0.923 for dice, IoU, and accuracy respectively. Compared to traditional segmentation models, the empirical results show that Tree-NET achieves higher accuracy with reduced training time and computational cost, thus representing a significant advancement in medical image analysis by providing more reliable and efficient tools for clinical applications.Master Thesis RNA Etkileşimlerinin İn Silico Analizi(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2024) Orhan, Mehmet Emin; Demirci, Müşerref Duygu SaçarMany supervised machine learning models have been developed for the classification and identification of non-coding RNA (ncRNA) sequences. These models play a significant role in the diagnosis and treatment of various diseases. During such analyses, positive learning datasets typically consist of known ncRNA examples, some of which may even be confirmed with strong experimental evidence. However, there is no database of validated negative sequences for ncRNA classes or standardized methodologies for generating high quality negative samples. To overcome this challenge, a new method for generating negative data called the NeRNA (Negative RNA) method has been developed in this study. NeRNA generates negative sequences using known ncRNA sequences and their octal representations, similar with frame shift mutations found in biology but without base deletions or insertions. In this thesis, the NeRNA method was tested separately with four different ncRNA datasets, including microRNA (miRNA), transfer RNA (tRNA), long non-coding RNA (lncRNA), and circular RNA (circRNA). Additionally, a species-specific case study was conducted to demonstrate and compare the performance of the study's miRNA predictions. The results of 1000-fold cross-validation on machine learning algorithms such as Decision Trees, Naive Bayes, Random Forest classifiers, and deep learning algorithms like Multilayer Perceptrons, Convolutional Neural Networks, and Simple Feedforward Neural Networks showed that models developed using datasets generated by NeRNA exhibited significantly high prediction performance. NeRNA has been published as an easy-to-use, updatable, and modifiable KNIME workflow, along with example datasets and required extensions that can be downloaded and utilized. NeRNA is designed specifically as a powerful tool for RNA sequence data analysis.Master Thesis Tasarlanmış Mikroorganizmalar ile Katma Değeri Yüksek Karotenoidlerin Biyosentezi(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2024) Arslansoy, Nuriye; Fidan, ÖzkanCarotenoids are pigment molecules that play an important role in coloring plants, algae, and other organisms. These molecules exhibit various biological activities such as anticancer, antiviral and antioxidant activities. They have a huge market size and are mainly used in the food, feed, and cosmetic industries. The current supply chain for carotenoids is mostly relied on the extraction from plants and/or chemical synthesis for certain carotenoids. However, these strategies have various bottlenecks and disadvantages such as being affected by climate change, more difficult and costly extraction processes, and environmental issues. These can be overcome with microbial biosynthesis, which not only addresses the previous problems but also provides advantages of producing in a short time and scale-up for industrial production. In this research, we aimed to biosynthesize the high value-added carotenoids by engineered microorganisms. The genome of a native producer of zeaxanthin diglucoside, identified as endophytic Pseudomonas sp. 102515, was first edited by CRISPR-Cas9 to knock out zeaxanthin glucosyltransferase (CrtX), lycopene β-cyclase (CrtY) and beta-carotene hydroxylase (CrtZ). This led to ΔcrtX, ΔcrtY and ΔcrtZ mutant strains of Pseudomonas sp. 102515. On the other hand, overexpression plasmids carrying crtW, CaZEP and CaZEP-CaCCSm40 genes were constructed and transformed to ΔcrtX mutant to synthesize astaxanthin, violaxanthin and capsanthin/capsorubin. HPLC analysis of extracts from mutant strains and overexpression strains revealed that all the engineered strains produced the corresponding carotenoids such as zeaxanthin, β-carotene, and lycopene. Thus, this study paved the way for the biosynthesis of valuable carotenoids in the engineered endophytic bacteria.Master Thesis Escherichia Coli Konak Organizmada GLP-1 Analoğunun Rekombinant Üretimi(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2024) Çalış, Burak; Fidan, ÖzkanDiabetes is the most serious metabolic disorder correlated with obesity, hypertension and cardiovascular conditions. High prevalence of Type II Diabetes Mellitus (T2DM) indicates the need for new medication development. In developing therapeutics, higher efficiency and fewer adverse effect features are targeted primarily. Recombinant protein-based biotechnological drug molecules have been developed and used for the treatment of T2DM. Especially, GLP-1 analogues are known by their self-limiting mechanism and insulinotropic effect. In this study, a novel GLP-1 analogue with increased stability and efficiency is produced using recombinant E. coli. The expression plasmid was constructed and confirmed by restriction digestion and whole plasmid sequencing. Then, itwas transformed into various E. coli strains followed by optimized lysis, growth and expression conditions to maximize the yield of the GLP-1 analogue. Various parameters such as pre-induction time, induction point, induction IPTG concentration and post-induction temperature were tested for the succesfull expression with maximum yield. Consequently, it was achieved that E. coli BL21(DE3) as strain, 0.2 mM IPTG induction at OD600nm of 0.6 and 18 °C overnight post-induction growth was the most promising conditions. Under these conditions, the GLP-1 analogue was obtained in the insoluble fraction. Following protein analysis and purification, quantification was performed and the highest titer of GLP-1 analogue was measured as 626 µg/ml. As future prospect, using another host organism and changing growth conditions can provide obtaining target protein in the soluble form. Keywords: T2DM, GLP-1 analogue, recombinant DNA technology, protein expression, E. coliMaster Thesis Potansiyel Gen Dağıtımı Uygulamaları için Poegma ve Sistaminle Modifiye Plazmit DNA'lar İçeren Polimerik Konjugatlar(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2024) Yıldız, Gizem; İşoğlu, İsmail Alper; İşoğlu, Sevil DinçerPolymer-based gene delivery systems have revealed significant advancements in the treatment of various diseases in recent years. Considering the potential of polymeric vectors, it is observed that the improvements in the field of gene therapy enable effective gene transfection and induced therapeutic protein production. In this thesis study, a strategy based on a new conjugation procedure is designed to increase the gene transfer and cellular uptake rate of plasmid DNAs. According to the findings, POEGMA-based carrier and cystamine-modified plasmid DNAs demonstrated successful conjugation through disulfide bond formation. MDA-MB-231 in vitro cellular uptake results of conjugates showed 94-98% cell internalization, indicating excellent results compared to the well-known polymers in the literature. As a result, the new delivery system we developed in this study determined the success of cystamine-modified plasmid DNAs binding to POEGMA polymer chains via a covalent linkage for the first time in the literature and provided a start for future studies.Master Thesis Peynirden İlk Defa İzole Edilen Loigolactobacillus Coryniformis FOL-19'un Yeni Nesil Dizilenmesi ve Diğer L. Coryniformis Suşlarıyla Karşılaştırmalı Genomik Analizleri(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2023) Gümüştop, İsmail; Ortakcı, FatihLoigolactobacillus coryniformis is a member of lactic acid bacteria isolated from various ecological niches. We isolated a novel L. coryniformis strain FOL-19 from artisanal Tulum cheese and performed the whole-genome sequencing for FOL-19 using Illumina NextSeq. Then, genomic characterization of FOL-19 against ten available whole genome sequences of the same species isolated from kimchi, silage, fermented meat, air of cowshed, and dairy was performed. The average genome size of 2.93 ±0.1 Mb, GC content of 42.96% ±0.002, number of CDS of 2905 ±165, number of tRNA of 56 ±10, and number of CRISPR elements of 6.55 ±1.83 was achieved. Both Type I and II Cas clusters were observed in L. coryniformis. Only one strain (CECT 5711) was predicted to encode a Carnocin CP52 bacteriocin gene cluster. The presence of CRISPR elements and Cas clusters suggests that L. coryniformis holds a promising potential for being a reservoir for new CRISPR-based tools. These findings put a step forward for the genomic characterization of L. coryniformis strains for biotechnological applications via genome-guided strain selection to identify industrially relevant traits.Master Thesis 32-mer MaSP1 Geninin pBbB6c Plazmid Vektörüne Klonlanması ve Escherichia Coli NEB 10-beta'ya Transformasyonu(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2023) Benk, Ruveyda; Ortakcı, Fatih; Öz, YahyaThe main purpose of my thesis was to clone Masp1 spider silk protein encoding gene from dragline type spider into E.coli NEB 10-beta organism. The recombinant microbial production of spider silk protein and converting it into a fiber format would ultimately produce a biomaterial also called as biosteel with high toughness and elasticity whereas low density compared to Kevlar, steel and carbon fiber. For this purpose, the gene encoding the dragline spider protein (MaSP1) was cloned into E. coli NEB 10-beta using recombinant molecular methods. First, 8-mer MaSP1 was synthesized and cloned via pGSI high copy cloning vector by sticky end cutting with restriction enzymes of KpnI,Kpn2I followed by heat-shock transformation into E.coli. Second, we performed restriction of the 8-mer plasmid by NheI and Kpn2I to extract the 8-mer. Later, the restriction was performed by SpeI and Kpn2I to obtain linearized pGSI containing 8-mer Masp1. A ligation was applied to merge 8-mer and pGSI plasmid carrying 8-mer Masp1 to achieve 16-mer Masp1 containing pGSI. Again, this plasmid was heat-shock transformed into E.coli. Following the same restriction 32-mer Masp1 containing pGSI plasmid was achieved. Finally, 32-mer Masp1 fragment was cut from pGSI cloning vector and ligated to pBbB6c low copy expression plasmid followed by electroporation into E.coli. The band size of 32-mer Masp1 gene was aligned between 3 kb and 5 kb which is an agreement with the calculated size of 32-mer Masp1 gene. Future studies should focus on the expression of Masp1 and the efficient production of this valuable recombinant protein under bioreactor conditions with cutting edge bioprocessing techniques.Master Thesis İn Siliko Analizlerle Yeni Patojenik Varyantlar Bulmak(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2022) Zorluer, Ziya Furkan; Kaplan, Oktay İsmailInherited diseases are health problems caused by one or more abnormalities in the genome. It can be caused by changes in a single gene (monogenic) or multiple genes (polygenic), or by a damage on chromosomes. Genetic variation is the differences in the DNA sequences that can be observed within a species or in alleles. Evaluation of genetic variants, together with reported phenotypic or pathogenic annotations from non-human organisms, facilitates the comparison of these variants with their human counterparts. In this work, we combined pathogenic and phenotypic annotations with variants, and these phenotypic orthologous variants from seven organisms can provide clues to the functional consequences of human genetic variants.Master Thesis Protein İkincil Yapı Tahmini için Boyut Küçültme(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2017) Görmez, Yasin; Aydın, Zafer; Kaynar, OğuzGerekli metabolik süreçleri yürüten proteinler insan hayatı için büyük önem taşımaktadır. Proteinlerin fonksiyonları ile üç boyutlu yapıları arasında yakın bir ilişki bulunmaktadır. Dört yapı düzeyi olan proteinlerin bir çoğunun, birincil yapı olarak da adlandırılan amino asit dizilimi bilinmekte ancak üçüncül yapıları bilinmemektedir. Üçüncül yapıların laboratuvar ortamında tespit edilmesinin çok maliyetli ve zor olması, amino asit dizilimini kullanarak yapı tahmini yapan sistemlerin geliştirilmesine neden olmuştur. Protein yapı tahmini yapan sistemlerin en önemli aşamalarından biri ise ikincil yapı etiketlerinin tanımlanması işlemidir. Yeni öznitelik çıkarma yaklaşımları geliştirildikçe yapısal özelliklerin tahmini için kullanılan veri setleri yüksek boyutlara sahip olabilmekte ve kullanılan özniteliklerden bazıları gürültülü veri içerebilmektedir. Bu nedenle uygun sayıda ve doğru öznitelikleri seçmek, iyi bir başarı oranı elde etmek için önemli aşamalardan biridir. Bu çalışmada iki farklı veri seti üzerinde derin oto kodlayıcı kullanılarak boyut düşürme işlemi uygulanmış, temel bileşen analizi, ki-kare, bilgi kazancı, kazanım oranı, korelasyon tabanlı öznitelik seçim teknikleri ve minimum fazlalık maksimum ilgi algoritması gibi çeşitli öznitelik seçim ve boyut düşürme teknikleri ayrıca genetik algoritma, aç gözlü algoritma ve en iyi ilk önce algoritması gibi çeşitli arama stratejileri ile birlikte kullanılarak elde edilen veri setleri ile karşılaştırılmıştır. İkincil yapı tahmin başarısının karşılaştırılması için destek vektör makinası kullanılmıştır.Master Thesis Cisplatin Temelli Nefrotoksisite Karşıtı Böbrek Hedefli Bir Nanotaşıyıcı Formülasyonu Geliştirilmesi(Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2019) Çakır, Şerife; Aydın, ErkinKitosan doğal bir polimer olup diğer sentetik polimerlere oranla vücutta daha az toksik etki göstermektedir. İyonik jelasyon metodu ile üretilen kitosan sodyum tripolifosfat (TPP) nanopartiküllerin böbrek ve beyin dokusu gibi insan vücut dokuları için iyi bir ilaç salınım araçları olduğu bilinmektedir. Bu çalışmada bir anti-kanser ilacı olan cisplatinin böbreklere oluşturduğu nefrotoksisiteyi gidermek için, gen susturucu siRNA'larla yüklü kitozan-TPP nanoparçacıkları kullanılmıştır. In vitro çalışmalar human kidney cell line olan Hek293 hücrelerinde denenmiş olup nanoparçacıkların hücreye girişleri ise floresan mikroskobu ve flow sitometri ile doğrulanmıştır. MTT ve XTT sonuçlarına göre nanoparçacıkların toksik etkisi düşük bulunmuştur. In vivo çalışmalara bakıldığında ise, balb-c tip 6-8 haftalık farelere siRNA yüklü nanoparçacık enjeksiyonu yapılmıştır. Sisplatin ile muamele edilmiş fareler kontrol ve siRNA-yüklü kitosan nanopartiküller grubu olarak hayvan grupları kullanılmıştır. Sisplatin enjeksiyonlarından sonra, siRNA-nanopartükül verilmesinden sonra farelerdeki kreatinin ve BUN seviyeleri değişimi incelendi. GAPDH bir kontrol geni olup PKC, P53, OCT1, OCT2 ve GGT genleri böbrek proximal tübül hücrelerinde önemli rollere sahiptir. Bu çalışmada bu genlerin mRNA seviyelerine de kantitatif PCR ile bakılmıştır. Enjeksiyonun ilk günlerinde siRNA'lar azalmış iken devam eden günlerde bu etki kaybolmuştur. Böylelikle her siRNA'nın susturma potansiyeli değişkenlik göstermektedir. Fakat bu değişkenlik çalışmada anlamlı bir değişim göstermektedir.