Modelling genetic variations with fragmentation-coagulation processes
We propose a novel class of Bayesian nonparametric models for sequential data called fragmentation-coagulation processes (FCPs). FCPs model a set of sequences using a partition-valued Markov process which evolves by splitting and merging clusters. An FCP is exchangeable, projective, stationary and r...
| Main Authors: | , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
2011
|
| _version_ | 1797087736775049216 |
|---|---|
| author | Teh, Y Blundell, C Elliott, LT |
| author_facet | Teh, Y Blundell, C Elliott, LT |
| author_sort | Teh, Y |
| collection | OXFORD |
| description | We propose a novel class of Bayesian nonparametric models for sequential data called fragmentation-coagulation processes (FCPs). FCPs model a set of sequences using a partition-valued Markov process which evolves by splitting and merging clusters. An FCP is exchangeable, projective, stationary and reversible, and its equilibrium distributions are given by the Chinese restaurant process. As opposed to hidden Markov models, FCPs allow for flexible modelling of the number of clusters, and they avoid label switching non-identifiability problems. We develop an efficient Gibbs sampler for FCPs which uses uniformization and the forward-backward algorithm. Our development of FCPs is motivated by applications in population genetics, and we demonstrate the utility of FCPs on problems of genotype imputation with phased and unphased SNP data. |
| first_indexed | 2024-03-07T02:39:56Z |
| format | Journal article |
| id | oxford-uuid:aa1409f2-644a-4e75-b3fd-6176cd9cbeaf |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T02:39:56Z |
| publishDate | 2011 |
| record_format | dspace |
| spelling | oxford-uuid:aa1409f2-644a-4e75-b3fd-6176cd9cbeaf2022-03-27T03:12:50ZModelling genetic variations with fragmentation-coagulation processesJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:aa1409f2-644a-4e75-b3fd-6176cd9cbeafEnglishSymplectic Elements at Oxford2011Teh, YBlundell, CElliott, LTWe propose a novel class of Bayesian nonparametric models for sequential data called fragmentation-coagulation processes (FCPs). FCPs model a set of sequences using a partition-valued Markov process which evolves by splitting and merging clusters. An FCP is exchangeable, projective, stationary and reversible, and its equilibrium distributions are given by the Chinese restaurant process. As opposed to hidden Markov models, FCPs allow for flexible modelling of the number of clusters, and they avoid label switching non-identifiability problems. We develop an efficient Gibbs sampler for FCPs which uses uniformization and the forward-backward algorithm. Our development of FCPs is motivated by applications in population genetics, and we demonstrate the utility of FCPs on problems of genotype imputation with phased and unphased SNP data. |
| spellingShingle | Teh, Y Blundell, C Elliott, LT Modelling genetic variations with fragmentation-coagulation processes |
| title | Modelling genetic variations with fragmentation-coagulation processes |
| title_full | Modelling genetic variations with fragmentation-coagulation processes |
| title_fullStr | Modelling genetic variations with fragmentation-coagulation processes |
| title_full_unstemmed | Modelling genetic variations with fragmentation-coagulation processes |
| title_short | Modelling genetic variations with fragmentation-coagulation processes |
| title_sort | modelling genetic variations with fragmentation coagulation processes |
| work_keys_str_mv | AT tehy modellinggeneticvariationswithfragmentationcoagulationprocesses AT blundellc modellinggeneticvariationswithfragmentationcoagulationprocesses AT elliottlt modellinggeneticvariationswithfragmentationcoagulationprocesses |