Single Cell Analysis of Z Ring Formation In Escherichia Coli Using Machine Learning Methods
Zare, Marzieh (2017)
Zare, Marzieh
2017
Master's Degree Programme in Information Technology
Tieto- ja sähkötekniikan tiedekunta - Faculty of Computing and Electrical Engineering
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2017-01-11
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201701021004
https://urn.fi/URN:NBN:fi:tty-201701021004
Tiivistelmä
In Escherichia coli (E. coli), Z-ring formation precedes the assembly of the membrane that partitions a cell into two daughter cells. As these FtsZ proteins are expressed, they first preferentially locate at the two cell poles. Afterwards, once the cell nucleoid splits in two and moves to the focal points of the cell, the FtsZ proteins start forming a ring at midcell. Finally, the ring becomes a circle, where the septum separating the nascent daughter cells forms.
In this thesis, we have used the most recent methods of image processing and machine learning to classify the stage of ring formation from microscopy images. We, first, segmented the cell images from microscopy images. Next, we performed the sample selection technique. After that, we extracted statistical features, i.e. mean and standard deviation (std), and then the samples were labelled by a biologist. We preferably applied three supervised classification methods, namely, Decision Tree (DT), Support Vector Machine (SVM), and Regularized Multinomial Logistic Regression (RMLR).
As a result, we found that RMLR performs better than two other classifiers. Accordingly, in the future we will use the RMLR algorithm to perform studies where the asymmetries arising from the stochasticity of FtsZ ring formations are analyzed.
In this thesis, we have used the most recent methods of image processing and machine learning to classify the stage of ring formation from microscopy images. We, first, segmented the cell images from microscopy images. Next, we performed the sample selection technique. After that, we extracted statistical features, i.e. mean and standard deviation (std), and then the samples were labelled by a biologist. We preferably applied three supervised classification methods, namely, Decision Tree (DT), Support Vector Machine (SVM), and Regularized Multinomial Logistic Regression (RMLR).
As a result, we found that RMLR performs better than two other classifiers. Accordingly, in the future we will use the RMLR algorithm to perform studies where the asymmetries arising from the stochasticity of FtsZ ring formations are analyzed.