Cluster analysis of AFLP data /
Amplified fragment length polymorphism (AFLP) is currently a widely used technique to study genome wide variation in closely related individuals. Unweighted paired group mean averaging (UPGMA) is a hierarchical clustering technique used to summarize the results of an AFLP study in a tree diagram ca...
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
2002.
|
| Subjects: | |
| Online Access: | http://proxy.library.tamu.edu/login?url=http://proquest.umi.com/pqdweb?did=765136581&sid=1&Fmt=2&clientId=2945&RQT=309&VName=PQD |
| Summary: | Amplified fragment length polymorphism (AFLP) is currently a widely used technique to study genome wide variation in closely related individuals. Unweighted paired group mean averaging (UPGMA) is a hierarchical clustering technique used to summarize the results of an AFLP study in a tree diagram called a dendrogram. UPGMA is an ad hoc technique that suffers from many shortcomings that can be addressed through statistical modeling. Some of these shortcomings include the lack of a readily interpretable measurement of uncertainty associated with the final estimate and a method for addressing measurement error. A Bayesian technique for clustering based on a special case of the beta-binomial distribution is developed. The technique provides a posterior probability for the uncertainty associated with the final clustering obtained while accounting for measurement errors. The technique is demonstrated on an AFLP data set of 11 varieties of grass. For this data set, the technique obtains clusterings with more similar clusters than other model based methods. |
|---|---|
| Item Description: | Vita. "Major Subject: Statistics". |
| Physical Description: | x, 76 leaves : illustrations ; 28 cm. Issued also on microfiche from University Microfilm Inc. |
| Bibliography: | Includes bibliographical references (leaves 70-71). |