2026-09-18T19:45:45Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/175362026-06-06T01:04:34Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023
Alan V. Oppenheim.
Boufounos, Petros T., 1977-
Massachusetts Institute of Technology. Dept. of Electrical Engineering and Computer Science.
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
2005-06-02T16:05:25Z
2005-06-02T16:05:25Z
2002
2002
http://hdl.handle.net/1721.1/17536
51111486
Thesis (M.Eng. and S.B.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2002.
Includes bibliographical references (p. 83-86).
DNA sequencing is the process of determining the sequence of chemical bases in a particular DNA molecule-nature's blueprint of how life works. The advancement of biological science in has created a vast demand for sequencing methods, which needs to be addressed by automated equipment. This thesis tries to address one part of that process, known as base calling: it is the conversion of the electrical signal-the electropherogram--collected by the sequencing equipment to a sequence of letters drawn from ( A,TC,G ) that corresponds to the sequence in the molecule sequenced. This work formulates the problem as a pattern recognition problem, and observes its striking resemblance to the speech recognition problem. We, therefore, propose combining Hidden Markov Models and Artificial Neural Networks to solve it. In the formulation we derive an algorithm for training both models together. Furthermore, we devise a method to create very accurate training data, requiring minimal hand-labeling. We compare our method with the de facto standard, PHRED, and produce comparable results. Finally, we propose alternative HMM topologies that have the potential to significantly improve the performance of the method.
by Petros T. Boufounos.
M.Eng.
S.B.
86 p.
3809782 bytes
3809587 bytes
application/pdf
application/pdf
eng
Massachusetts Institute of Technology
M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission.
http://dspace.mit.edu/handle/1721.1/7582
Electrical Engineering and Computer Science.
Nucleotide sequence
Signal processing for DNA sequencing
Signal processing for Deoxyribonucleic acid sequencing
Thesis
application/pdf
Publication
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="b56706a7-c351-4ac3-bd18-a736f1076bed">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Language>eng</Language>
<Title>Signal processing for DNA sequencing</Title>
<Subtitle>Signal processing for Deoxyribonucleic acid sequencing</Subtitle>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2002</PublicationDate>
<Authors>
<Author>
<DisplayName>Boufounos, Petros T., 1977-</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>http://dspace.mit.edu/handle/1721.1/7582</License>
<Keyword>Electrical Engineering and Computer Science.</Keyword>
<Abstract>DNA sequencing is the process of determining the sequence of chemical bases in a particular DNA molecule-nature's blueprint of how life works. The advancement of biological science in has created a vast demand for sequencing methods, which needs to be addressed by automated equipment. This thesis tries to address one part of that process, known as base calling: it is the conversion of the electrical signal-the electropherogram--collected by the sequencing equipment to a sequence of letters drawn from ( A,TC,G ) that corresponds to the sequence in the molecule sequenced. This work formulates the problem as a pattern recognition problem, and observes its striking resemblance to the speech recognition problem. We, therefore, propose combining Hidden Markov Models and Artificial Neural Networks to solve it. In the formulation we derive an algorithm for training both models together. Furthermore, we devise a method to create very accurate training data, requiring minimal hand-labeling. We compare our method with the de facto standard, PHRED, and produce comparable results. Finally, we propose alternative HMM topologies that have the potential to significantly improve the performance of the method.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>