Researcher:
Bozkurt, Yasemin

Profile Picture
ORCID

Job Title

PhD Student

First Name

Yasemin

Last Name

Bozkurt

Name

Name Variants

Bozkurt, Yasemin

Email Address

Birth Date

Filters

2015 - 20203

Advanced Search

Filter by

Settings

Researcher Of Publications: search.filters.isAuthorOfPublication.7efae058-ed3a-45b8-bed0-52c2efbfd18e

Search Results

Now showing 1 - 3 of 3
  • Placeholder
    PublicationMetadata only
    Interpretable embeddings from molecular simulations using Gaussian mixture variational autoencoders
    (IOP Publishing Ltd, 2020) Bereau, Tristan; Rudzinski, Joseph F.; Bozkurt, Yasemin; PhD Student; Graduate School of Sciences and Engineering; N/A
    Extracting insight from the enormous quantity of data generated from molecular simulations requires the identification of a small number of collective variables whose corresponding low-dimensional free-energy landscape retains the essential features of the underlying system. Data-driven techniques provide a systematic route to constructing this landscape, without the need for extensive a priori intuition into the relevant driving forces. In particular, autoencoders are powerful tools for dimensionality reduction, as they naturally force an information bottleneck and, thereby, a low-dimensional embedding of the essential features. While variational autoencoders ensure continuity of the embedding by assuming a unimodal Gaussian prior, this is at odds with the multi-basin free-energy landscapes that typically arise from the identification of meaningful collective variables. In this work, we incorporate this physical intuition into the prior by employing a Gaussian mixture variational autoencoder (GMVAE), which encourages the separation of metastable states within the embedding. The GMVAE performs dimensionality reduction and clustering within a single unified framework, and is capable of identifying the inherent dimensionality of the input data, in terms of the number of Gaussians required to categorize the data. We illustrate our approach on two toy models, alanine dipeptide, and a challenging disordered peptide ensemble, demonstrating the enhanced clustering effect of the GMVAE prior compared to standard VAEs. The resulting embeddings appear to be promising representations for constructing Markov state models, highlighting the transferability of the dimensionality reduction from static equilibrium properties to dynamics.
  • Placeholder
    PublicationMetadata only
    Mathematical modeling of Behçet's disease: a dynamical systems approach
    (World Scientific Publ Co Pte Ltd, 2015) Gül, Ahmet; N/A; N/A; Department of Chemical and Biological Engineering; Department of Electrical and Electronics Engineering; Erdem, Cemal; Bozkurt, Yasemin; Erman, Burak; Demir, Alper; Master Student; PhD Student; Faculty Member; Faculty Member; Department of Chemical and Biological Engineering; Department of Electrical and Electronics Engineering; N/A; Graduate School of Sciences and Engineering; College of Engineering; College of Engineering; N/A; N/A; 179997; 3756
    Behcet's Disease (BD) is a multi-systemic, auto-inflammatory disorder that is characterized by recurrent episodes of inflammatory manifestations affecting skin, mucosa, eyes, blood vessels, joints and several other organs. BD is classified as a multifactorial disease with an important contribution of genetics. Genetic studies suggest that there is a strong association of BD with a Class I major histocompatibility complex antigen, named HLA-B*51, along with several other weaker associations with genes encoding proteins involved in inflammation. However, pathogenic mechanisms associated with these genetic variations and their interactions with the environment have not been elucidated yet. In this paper, we present a mathematical model for BD based on a dynamical systems perspective that captures especially the relapsing nature of the disease. We propose a disease progression mechanism and construct a model, in the form of coupled ordinary differential equations (ODEs), which reveals the occurrence pattern of the disease in the population. According to our model, the disease has three distinct modes describing different phenotypes of people carrying HLA-B*51 tissue antigen, namely, the Healthy Carrier, the Potential Patient and the Active Patient. We herein present an exemplary mathematical model for BD, for the first time in the literature, that concisely captures the actions of many cell types together with genetic and environmental effects. The proposed model provides insight into this complex inflammatory disease which may lead to identification of new tools for its treatment and prevention.
  • Thumbnail Image
    PublicationOpen Access
    Unified modeling of familial mediterranean fever and cryopyrin associated periodic syndromes
    (BioMed Central, 2015) Gul, Ahmet; N/A; Department of Electrical and Electronics Engineering; Department of Chemical and Biological Engineering; Bozkurt, Yasemin; Demir, Alper; Erman, Burak; Master Student - PhD Student; Faculty Member; Faculty Member; Department of Electrical and Electronics Engineering; Department of Chemical and Biological Engineering; Graduate School of Sciences and Engineering; College of Engineering; N/A; 3756; 179997
    Familial mediterranean fever (FMF) and Cryopyrin associated periodic syndromes (CAPS) are two prototypical hereditary autoinflammatory diseases, characterized by recurrent episodes of fever and inflammation as a result of mutations in MEFV and NLRP3 genes encoding Pyrin and Cryopyrin proteins, respectively. Pyrin and Cryopyrin play key roles in the multiprotein inflammasome complex assembly, which regulates activity of an enzyme, Caspase 1, and its target cytokine, IL-1 beta. Overproduction of IL-1 beta by Caspase 1 is the main cause of episodic fever and inflammatory findings in FMF and CAPS. We present a unifying dynamical model for FMF and CAPS in the form of coupled nonlinear ordinary differential equations. The model is composed of two subsystems, which capture the interactions and dynamics of the key molecular players and the insults on the immune system. One of the subsystems, which contains a coupled positive-negative feedback motif, captures the dynamics of inflammation formation and regulation. We perform a comprehensive bifurcation analysis of the model and show that it exhibits three modes, capturing the Healthy, FMF, and CAPS cases. The mutations in Pyrin and Cryopyrin are reflected in the values of three parameters in the model. We present extensive simulation results for the model that match clinical observations.