Collective Estimation of Multiple Bivariate Density Functions With Application to Angular-Sampling-Based Protein Loop Modeling
Maadooliat, Mehdi ; Zhou, Lan ; Najibi, Seyed Morteza LU
; Gao, Xin
and Huang, Jianhua Z.
(2016)
In Journal of the American Statistical Association
111(513).
p.43-56
- Abstract
This article develops a method for simultaneous estimation of density functions for a collection of populations of protein backbone angle pairs using a data-driven, shared basis that is constructed by bivariate spline functions defined on a triangulation of the bivariate domain. The circular nature of angular data is taken into account by imposing appropriate smoothness constraints across boundaries of the triangles. Maximum penalized likelihood is used to fit the model and an alternating blockwise Newton-type algorithm is developed for computation. A simulation study shows that the collective estimation approach is statistically more efficient than estimating the densities individually. The proposed method was used to estimate... (More)
This article develops a method for simultaneous estimation of density functions for a collection of populations of protein backbone angle pairs using a data-driven, shared basis that is constructed by bivariate spline functions defined on a triangulation of the bivariate domain. The circular nature of angular data is taken into account by imposing appropriate smoothness constraints across boundaries of the triangles. Maximum penalized likelihood is used to fit the model and an alternating blockwise Newton-type algorithm is developed for computation. A simulation study shows that the collective estimation approach is statistically more efficient than estimating the densities individually. The proposed method was used to estimate neighbor-dependent distributions of protein backbone dihedral angles (i.e., Ramachandran distributions). The estimated distributions were applied to protein loop modeling, one of the most challenging open problems in protein structure prediction, by feeding them into an angular-sampling-based loop structure prediction framework. Our estimated distributions compared favorably to the Ramachandran distributions estimated by fitting a hierarchical Dirichlet process model; and in particular, our distributions showed significant improvements on the hard cases where existing methods do not work well.
(Less)
- author
- Maadooliat, Mehdi
; Zhou, Lan
; Najibi, Seyed Morteza
LU
; Gao, Xin
and Huang, Jianhua Z.
- publishing date
- 2016-01-02
- type
- Contribution to journal
- publication status
- published
- keywords
- Bivariate splines, Log-spline density estimation, Protein structure, Ramachandran distribution, Roughness penalty, Triangulations
- in
- Journal of the American Statistical Association
- volume
- 111
- issue
- 513
- pages
- 14 pages
- publisher
- American Statistical Association
- external identifiers
-
- scopus:84969750832
- ISSN
- 0162-1459
- DOI
- 10.1080/01621459.2015.1099535
- language
- English
- LU publication?
- no
- id
- e45a4e93-6728-485a-9dc6-1b6dc7fac2c7
- date added to LUP
- 2019-09-23 11:21:39
- date last changed
- 2022-02-16 00:09:40
@article{e45a4e93-6728-485a-9dc6-1b6dc7fac2c7,
abstract = {{<p>This article develops a method for simultaneous estimation of density functions for a collection of populations of protein backbone angle pairs using a data-driven, shared basis that is constructed by bivariate spline functions defined on a triangulation of the bivariate domain. The circular nature of angular data is taken into account by imposing appropriate smoothness constraints across boundaries of the triangles. Maximum penalized likelihood is used to fit the model and an alternating blockwise Newton-type algorithm is developed for computation. A simulation study shows that the collective estimation approach is statistically more efficient than estimating the densities individually. The proposed method was used to estimate neighbor-dependent distributions of protein backbone dihedral angles (i.e., Ramachandran distributions). The estimated distributions were applied to protein loop modeling, one of the most challenging open problems in protein structure prediction, by feeding them into an angular-sampling-based loop structure prediction framework. Our estimated distributions compared favorably to the Ramachandran distributions estimated by fitting a hierarchical Dirichlet process model; and in particular, our distributions showed significant improvements on the hard cases where existing methods do not work well.</p>}},
author = {{Maadooliat, Mehdi and Zhou, Lan and Najibi, Seyed Morteza and Gao, Xin and Huang, Jianhua Z.}},
issn = {{0162-1459}},
keywords = {{Bivariate splines; Log-spline density estimation; Protein structure; Ramachandran distribution; Roughness penalty; Triangulations}},
language = {{eng}},
month = {{01}},
number = {{513}},
pages = {{43--56}},
publisher = {{American Statistical Association}},
series = {{Journal of the American Statistical Association}},
title = {{Collective Estimation of Multiple Bivariate Density Functions With Application to Angular-Sampling-Based Protein Loop Modeling}},
url = {{http://dx.doi.org/10.1080/01621459.2015.1099535}},
doi = {{10.1080/01621459.2015.1099535}},
volume = {{111}},
year = {{2016}},
}