Epistasis: Methods and Protocols

352 Pages·2015·10.835 MB·English

by  Jason H. Moore, Scott M. Williams (eds.)

Methods in
Molecular Biology 1253
Jason H. Moore
Scott M. Williams Editors
Epistasis
Methods and Protocols
M M B
ETHODS IN OLECULAR IOLOGY
Series Editor
John M. Walker
School of Life Sciences
University of Hertfordshire
Hat fi eld, Hertfordshire, AL10 9AB, UK
For further volumes:
http://www.springer.com/series/7651

Epistasis
Methods and Protocols
Edited by
Jason H. Moore
Department of Genetics, Institute for Quantitative Biomedical Sciences, Geisel School of Medicine,
Hanover, NH, USA
Scott M. Williams
Department of Genetics, Institute of Quantitative Biomedical Sciences, Geisel School of Medicine,
Hanover, NH, USA
Editors
Jason H. M oore Scott M. W illiams
Department of Genetics Department of Genetics
Institute for Quantitative Biomedical Sciences Institute of Quantitative Biomedical Sciences
Geisel School of Medicine Geisel School of Medicine
Hanover, NH , USA Hanover , NH, USA
ISSN 1064-3745 ISSN 1940-6029 (electronic)
ISBN 978-1-4939-2154-6 ISBN 978-1-4939-2155-3 (eBook)
DOI 10.1007/978-1-4939-2155-3
Springer New York Heidelberg Dordrecht London
Library of Congress Control Number: 2014953909
© Springer Science+Business Media New York 2 015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifi cally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfi lms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this
legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifi cally for
the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions
for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution
under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specifi c statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication, neither
the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be
made. The publisher makes no warranty, express or implied, with respect to the material contained herein.
Printed on acid-free paper
Humana Press is a brand of Springer
Springer is part of Springer Science+Business Media (www.springer.com)
Prefa ce
Modern genetic analyses, with phenomenal technological advances, now permit deeper
interrogation of genomes with the intent of constructing more accurate and comprehensive
genotype to phenotype maps. However, as recognized by the authors of the chapters in this
volume, a key to defi ning this map requires inclusion of factors not always explicitly incor-
porated into genetic analyses—namely, epistasis or interactions. Not doing this has led, at
least in part, to less than perfect descriptions of genotype to phenotype maps and has moti-
vated the term, missing heritability, or the amount of genetic variance of a trait that is left
unexplained. The key is that even with enormous quantities of data, fully explanatory
genetic models evade description, if inappropriate simplifying assumptions are made. The
focus of this book is to explore how we can avoid making these assumptions and do so in
ways that are practical.
One key to unraveling the role of epistasis in genotype to phenotype maps is to mini-
mize the extraordinary number of possible interactions that can be assessed in genome-
wide data sets, hence predefi ning the set of possible models of epistasis that are to be
included in analyses. Such fi ltering can serve as a precursor to statistical or data mining
analyses, both of which are covered in this book. With respect to appropriate statistical
analyses for the detection of epistasis, it is important to precisely defi ne the meaning of
epistasis to be included in analyses, as historically more than one defi nition has existed, and
they can create ambiguities in terms of how epistasis is tested. Therefore, several of our
authors take substantial space to defi ne epistasis with respect to how to appropriately ana-
lyze it. Lastly, genomic data can be mined using a variety of computational tools that make
no a priori assumptions about the underlying genetic models. These are promising but
often make interpretation diffi cult.
As any genotype to phenotype map is determined by the history of the genome in ques-
tion, it is important to defi ne how evolutionary processes may have shaped a trait’s genetic
architecture. This is addressed in Chapter 1 . Methods that reduce the multiple testing bur-
den are described in Chapters 2 and 3 . An alternative approach in model systems is to per-
turb the “natural” genetic system by generating de novo mutations and assessing their roles
via quantitative trait locus mapping in multiple backgrounds (Chapter 4 ) . In systems not
amenable to such manipulation (e.g., humans) epistasis analyses may depend on well-
chosen candidates, an approach shown to work for neuropsychiatric diseases where epistasis
and pleiotropy appear to overlap (Chapter 5 ) .
In Chapter 6 the authors discuss the decomposition of genetic variance into its individ-
ual components, how this underpins our understanding of epistasis, and how this may affect
the outcome of selection. Measuring epistasis is a key topic of Chapter 7 , where it is argued
that how epistasis is measured can appear to minimize its effects in an evolutionary context.
The role of measurement of epistasis is also taken up in Chapter 8 , where it is shown that the
arbitrariness of epistasis or interaction can be eliminated by applying measurement theoretic
constraints. Extending the allelic average excess and average effect to two or more loci is
proposed as a novel analytical approach in Chapter 9 . By explicitly d efi ning capacitating
epistasis in Chapter 1 0 , the authors develop means to examine its effects.
v
vi Preface
Distinct from most other chapters, the authors of Chapter 1 1 take an explicitly epide-
miological view of what they defi ne as “compositional” epistasis, and how to best detect it.
Chapter 12 examines Boolean function interactions in Age-related Macular Degeneration
data and fi nds relevant gene–gene and gene–environment interactions. Using information
theory to detect and characterize epistasis is the focus of Chapter 1 3 . Chapters 1 4 and 1 5
examine the application of network building to better elucidating epistasis. Agnostic data
mining methods are the core of Chapters 16 and 1 7 where two methods, multifactor
dimensionality reduction and ReliefF, are described. Lastly, artifi cial intelligence methods
are introduced in Chapter 1 8 as a means to detect epistasis in association studies.
Overall, we think that the chapters provide a comprehensive set of ideas that can help
us elucidate epistasis in the context of modern data availability, and thereby help us to bet-
ter understand the genetic bases of complex phenotypes and their evolutionary histories.
Hanover, NH, USA Jason H. M oore
S cott M. Williams
Contents
Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i x
1 Long-Term Selection Experiments: Epistasis and the Response to Selection. . . 1
Charles Goodnight
2 Finding the Epistasis Needles in the Genome-Wide Haystack . . . . . . . . . . . . . 1 9
Marylyn D. Ritchie
3 B iological Knowledge-Driven Analysis of Epistasis in Human GWAS
with Application to Lipid Traits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Li Ma , A lon K einan , and Andrew G . C lark
4 E pistasis for Quantitative Traits in Drosophila . . . . . . . . . . . . . . . . . . . . . . . . . 4 7
Trudy F. C . Mackay
5 E pistasis in the Risk of Human Neuropsychiatric Disease. . . . . . . . . . . . . . . . . 7 1
Scott M. W illiams
6 O n the Partitioning of Genetic Variance with Epistasis . . . . . . . . . . . . . . . . . . 9 5
José M. Á lvarez-Castro and Arnaud L e Rouzic
7 M easuring Gene Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 15
Thomas F . H ansen
8 T wo Rules for the Detection and Quantification of Epistasis
and Other Interaction Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Günter P . W agner
9 D irect Approach to Modeling Epistasis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 59
Rong-Cai Y ang
10 C apacitating Epistasis—Detection and Role in the Genetic Architecture
of Complex Traits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Mats E . P ettersson and Örjan Carlborg
11 C ompositional Epistasis: An Epidemiologic Perspective. . . . . . . . . . . . . . . . . . 1 97
Etsuji S uzuki and Tyler J. V anderWeele
12 Identification of Genome-Wide SNP–SNP and SNP–Clinical Boolean
Interactions in Age-Related Macular Degeneration . . . . . . . . . . . . . . . . . . . . . 2 17
Carlos R iveros , Renato Vimieiro , E lizabeth G . Holliday ,
Christopher Oldmeadow , Jie J in Wang , Paul Mitchell , John A ttia ,
Rodney J . S cott , and Pablo A. M oscato
13 Epistasis Analysis Using Information Theory. . . . . . . . . . . . . . . . . . . . . . . . . . 2 57
Jason H. Moore and Ting H u
14 G enome-Wide Epistasis and Pleiotropy Characterized
by the Bipartite Human Phenotype Network. . . . . . . . . . . . . . . . . . . . . . . . . . 269
Christian Darabos and Jason H . M oore
vii
viii Contents
15 Network Theory for Data-Driven Epistasis Networks . . . . . . . . . . . . . . . . . . . 2 85
Caleb A . L areau and Brett A . M cKinney
16 E pistasis Analysis Using Multifactor Dimensionality Reduction . . . . . . . . . . . . 3 01
Jason H . Moore and Peter C . A ndrews
17 E pistasis Analysis Using ReliefF. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Jason H. M oore
18 Epistasis Analysis Using Artificial Intelligence . . . . . . . . . . . . . . . . . . . . . . . . . 3 27
Jason H . Moore and Doug P . H ill
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 47
Contributors
JOSÉ M. Á LVAREZ-CASTRO • Department of Genetics, U niversity of Santiago de Compostela ,
Lugo , G aliza, S pain ; I nstituto Gulbenkian de Ciência , Oeiras , P ortugal
PETER C. ANDREWS • Department of Genetics, G eisel School of Medicine, DHMC , Lebanon,
NH, U SA
JON ATTIA • School of Medicine and Public Health, The University of Newcastle , C allaghan,
NSW , A ustralia ; CREDITSS – Clinical Research Design, Information Technology and
Statistical Support Unit, Hunter Medical Research Institute , N ew Lambton Heights,
NSW , A ustralia
ÖRJAN CARLBORG • Division of Computational Genetics, Department of Clinical Sciences ,
Swedish University of Agricultural Sciences , U ppsala , Sweden
ANDREW G . CLARK • Department of Biological Statistics and Computational Biology,
Cornell University , I thaca , N Y , USA ; Department of Molecular Biology and Genetics,
Cornell University , Ithaca, NY , U SA
CHRISTIAN D ARABOS • Department of Genetics, Geisel School of Medicine, DHMC , L ebanon,
NH, U SA
CHARLES G OODNIGHT • Department of Biology, University of Vermont , Burlington, VT, USA
THOMAS F. H ANSEN • Department of Biology, Centre for Ecological and Evolutionary
Synthesis, U niversity of Oslo , Blindern, N orway
DOUG P . HILL • Department of Genetics, G eisel School of Medicine, DHMC , L ebanon, N H ,
USA
ELIZABETH G. HOLLIDAY • School of Medicine and Public Health, The University
of Newcastle , C allaghan, N SW , Australia ; CREDITSS – Clinical Research Design,
Information Technology and Statistical Support Unit , Hunter Medical Research
Institute , New Lambton Heights, N SW , Australia
TING HU • Department of Genetics, G eisel School of Medicine, DHMC , Lebanon, NH, U SA
ALON KEINAN • Department of Biological Statistics and Computational Biology, Cornell
Center for Comparative and Population Genomics, C ornell University , I thaca , N Y , U SA
CALEB A. LAREAU • Department of Mathematics, U niversity of Tulsa , T ulsa , O K , U SA
LI MA • Department of Animal and Avian Sciences , University of Maryland , College Park,
MD, U SA
TRUDY F. C. M ACKAY • Department of Biological Sciences, N orth Carolina State University ,
Raleigh, N C , U SA
BRETT. A. MCKINNEY • Department of Mathematics, Tandy School of Computer Science,
Laureate Institute for Brain Research , University of Tulsa , Tulsa, O K , U SA
PAUL M ITCHELL • Department of Ophthalmology, Westmead Centre for Vision Research,
Millennium Institute, University of Sydney , Sydney, A ustralia
JASON H. M OORE • Department of Genetics, I nstitute for Quantitative Biomedical Sciences,
Geisel School of Medicine, Hanover , N H , U SA ; D epartment of Community and Family
Medicine, Geisel School of Medicine, DHMC , L ebanon, N H , U SA
ix