Download Gene Mapping for Correlated Traits

Gene Mapping for Correlated Traits Brian S. Yandell University of Wisconsin-­‐Madison www.stat.wisc.edu/~yandell/statgen Correlated Traits UCLA Networks (c) Yandell 2013 1 www.nobelprize.org/educational/medicine/dna
SysGen: Overview
www.accessexcellence.org/RC/VL/GG/central.php
Seattle SISG: Yandell © 2012
2
phenotypic buffering
of molecular QTL
Fu et al. Jansen (2009 Nature Genetics)
SysGen: Overview
Seattle SISG: Yandell © 2012
3
Biochemical Pathways chart, Gerhard Michal, Beohringer Mannheim
http://web.expasy.org/pathways/
SysGen: Overview
Seattle SISG: Yandell © 2012
4
http://web.expasy.org/pathways/
SysGen: Overview
Seattle SISG: Yandell © 2012
5
KEGG pathway: pparg in mouse
SysGen: Overview
Seattle SISG: Yandell © 2012
6
systems genetics approach
•  study geneKc architecture of quanKtaKve traits –  in model systems, and ulKmately humans •  interrogate single resource populaKon for variaKon –  DNA sequence, transcript abundance, proteins, metabolites –  mulKple organismal phenotypes –  mulKple environments •  detailed map of geneKc variants associated with –  each organismal phenotype in each environment •  funcKonal context to interpret phenotypes –  geneKc underpinnings of mulKple phenotypes –  geneKc basis of genotype by environment interacKon Sieberts, Schadt (2007 Mamm Genome); Emilsson et al. (2008 Nature) Chen et al. 2008 Nature); Ayroles et al. MacKay (2009 Nature Gene-cs) SysGen: Overview
Seattle SISG: Yandell © 2012
7
brief tutorial on gene mapping for experimental crosses *** Check out Karl Broman’s stuff *** –  visit www.rqtl.org –  go through talk on mulKple QTLs (for instance) h]p://www.biostat.wisc.edu/~kbroman/
presentaKons/mulKqtl_columbia11.pdf –  Work through tutorials at www.rqtl.org Correlated Traits UCLA Networks (c) Yandell 2013 8 eQTL Tools
Seattle SISG: Yandell © 2010
9
Gene-c architecture of gene expression in 6 -ssues. A Tissue-­‐specific panels illustrate the rela-onship between the genomic loca-on of a gene (y-­‐axis) to where that gene’s mRNA shows an eQTL (LOD > 5), as a func-on of genome posi-on (x-­‐axis). Circles represent eQTLs that showed either cis-­‐linkage (black) or trans-­‐linkage (colored) according to LOD score. Genomic hot spots, where many eQTLs map in trans, are apparent as ver-cal bands that show either -ssue selec-vity (e.g., Chr 6 in the islet, s) or are present in all -ssues (e.g., Chr 17, q). B The total number of eQTLs iden-fied in 5 cM genomic windows is ploPed for each -ssue; total eQTLs for all posi-ons is shown in upper right corner for each panel. The peak number of eQTLs exceeding 1000 per 5 cM is shown for islets (Chrs 2, 6 and 17), liver (Chrs 2 and 17) and kidney (Chr 17). Figure 4 Tissue-­‐specific hotspots with eQTL and SNP architecture for Chrs 1, 2 and 17. The number of eQTLs for each -ssue (leV axis) and the number of SNPs between B6 and BTBR (right axis) that were iden-fied within a 5 cM genomic window is shown for Chr 1 (A), Chr 2 (B) Chr 17 (C). The loca-on of -ssue-­‐specific hotspots are iden-fied by their number corresponding to that in Table 1. eQTL and SNP architecture is shown for all chromosomes in supplementary material. BxH ApoE-/- chr 2: causal architecture
hotspot
12 causal calls
eQTL Tools
Seattle SISG: Yandell © 2010
12
BxH ApoE-­‐/-­‐ causal network for transcripKon factor Pscdbp causal trait
work of
Elias Chaibub Neto
eQTL Tools
Seattle SISG: Yandell © 2010
13
MulKple Correlated Traits •  Pleiotropy vs. close linkage •  Analysis of covariance –  Regress one trait on another before QTL search • 
• 
• 
• 
Classic GxE analysis Formal joint mapping (MTM) Seemingly unrelated regression (SUR) Reducing many traits to one –  Principle components for similar traits Correlated Traits UCLA Networks (c) Yandell 2013 14 co-­‐mapping mulKple traits •  avoid reducKonist approach to biology –  address physiological/biochemical mechanisms –  Schmalhausen (1942); Falconer (1952) •  separate close linkage from pleiotropy –  1 locus or 2 linked loci? •  idenKfy epistaKc interacKon or canalizaKon –  influence of geneKc background •  establish QTL x environment interacKons •  decompose geneKc correlaKon among traits •  increase power to detect QTL Correlated Traits UCLA Networks (c) Yandell 2013 15 Two types of data •  Design I: mulKple traits on same individual –  Related measurements, say of shape or size –  Same measurement taken over Kme –  CorrelaKon within an individual •  Design II: mulKple traits on different individuals –  Same measurement in two crosses –  Male vs. female differences –  Different individuals in different locaKons –  No correlaKon between individuals Correlated Traits UCLA Networks (c) Yandell 2013 16 interplay of pleiotropy & correlaKon pleiotropy only correlaKon only both Korol et al. (2001) Correlated Traits UCLA Networks (c) Yandell 2013 17 Brassica napus: 2 correlated traits •  4-­‐week & 8-­‐week vernalizaKon effect –  log(days to flower) •  geneKc cross of –  Stellar (annual canola) –  Major (biennial rapeseed) •  105 F1-­‐derived double haploid (DH) lines –  homozygous at every locus (QQ or qq) •  10 molecular markers (RFLPs) on LG9 –  two QTLs inferred on LG9 (now chromosome N2) –  corroborated by Butruille (1998) –  exploiKng synteny with Arabidopsis thaliana Correlated Traits UCLA Networks (c) Yandell 2013 18 QTL with GxE or Covariates • 
adjust phenotype by covariate –  covariate(s) = environment(s) or other trait(s) • 
addiKve covariate –  covariate adjustment same across genotypes –  “usual” analysis of covariance (ANCOVA) • 
interacKng covariate –  address GxE –  capture genotype-­‐specific relaKonship among traits • 
another way to think of mulKple trait analysis –  examine single phenotype adjusted for others Correlated Traits UCLA Networks (c) Yandell 2013 19 Correlated Traits UCLA Networks (c) Yandell 2013 20 Correlated Traits UCLA Networks (c) Yandell 2013 21 Correlated Traits UCLA Networks (c) Yandell 2013 22 MulKple trait mapping •  Joint mapping of QTL –  tesKng and esKmaKng QTL affecKng mulKple traits •  TesKng pleiotropy vs. close linkage –  One QTL or two closely linked QTLs •  TesKng QTL x environment interacKon •  Comprehensive model of mulKple traits –  Separate geneKc & environmental correlaKon Correlated Traits UCLA Networks (c) Yandell 2013 23 3 correlated traits (Jiang Zeng 1995) 0
jiang2
1
2
3
ρP = 0.06, ρG = 0.68, ρE = -0.2
-3
-2
-1
ellipses centered on genotypic value width for nominal frequency main axis angle environmental correlaKon 3 QTL, F2 27 genotypes ρP = 0.3, ρG = 0.54, ρE = 0.2
note signs of geneKc and environmental correlaKon -2
-1
0
1
2
jiang3
ρP = -0.07, ρG = -0.22, ρE = 0
3
0
jiang1
-2
-2
-1
-1
0
jiang1
1
1
2
2
-3
-3
Correlated Traits -2
-1
0
jiang2
1
2
UCLA Networks (c) Yandell 2013 3
-2
-1
0
jiang3
1
2
24 pleiotropy or close linkage? 2 traits, 2 qtl/trait pleiotropy @ 54cM linkage @ 114,128cM Jiang Zeng (1995) Correlated Traits UCLA Networks (c) Yandell 2013 25 Formal Tests: 2 traits y1 ~ N(μq1, σ2) for group 1 with QTL at locaKon λ1 y2 ~ N(μq2, σ2) for group 2 with QTL at locaKon λ2 •  Pleiotropy vs. close linkage •  test QTL at same locaKon: λ1 = λ2 •  likelihood raKo test (LOD): null forces same locaKon •  if pleiotropic (λ1 = λ2) •  test for same mean: μq1 = μq2 •  Likelihood raKo test (LOD) •  null forces same mean, locaKon •  alternaKve forces same locaKon •  only make sense if traits are on same scale •  test sex or locaKon effect Correlated Traits UCLA Networks (c) Yandell 2013 26 More detail for 2 traits y1 ~ N(μq1, σ2) for group 1 y2 ~ N(μq2, σ2) for group 2 •  two possible QTLs at locaKons λ1 and λ2 •  effect βkj in group k for QTL at locaKon λj μq1 = μ1 + β11(q1) + β12(q2) μq2 = μ2 + β21(q1) + β22(q2) •  classical: test βkj = 0 for various combinaKons Correlated Traits UCLA Networks (c) Yandell 2013 27 reducing many phenotypes to 1 •  Drosophila mauri-ana x D. simulans –  reciprocal backcrosses, ~500 per bc •  response is “shape” of reproducKve piece –  trace edge, convert to Fourier series –  reduce dimension: first principal component •  many linked loci –  brief comparison of CIM, MIM, BIM Correlated Traits UCLA Networks (c) Yandell 2013 28 7.6
-0.2
7.8
-0.1
8.0
ettf1
8.2
PC2 (7%)
0.0
0.1
8.4
0.2
8.6
PC for two correlated phenotypes 8.2
8.4
Correlated Traits 8.6
8.8
9.0
etif3s6
9.2
9.4
UCLA Networks (c) Yandell 2013 -0.5
0.0
PC1 (93%)
0.5
29 shape phenotype via PC Liu et al. (1996) Gene-cs Correlated Traits UCLA Networks (c) Yandell 2013 30 shape phenotype in BC study indexed by PC1 Liu et al. (1996) Gene-cs Correlated Traits UCLA Networks (c) Yandell 2013 31 Zeng et al. (2000) CIM vs. MIM composite interval mapping (Liu et al. 1996) narrow peaks miss some QTL mulKple interval mapping (Zeng et al. 2000) triangular peaks both condiKonal 1-­‐D scans fixing all other "QTL" Correlated Traits UCLA Networks (c) Yandell 2013 32 mulKple QTL: CIM, MIM and BIM cim bim mim Correlated Traits UCLA Networks (c) Yandell 2013 33 MulK-­‐trait strategy •  Use some data-­‐reducKon method –  Principal components or clustering –  WGCNA modules –  Gene-­‐mapping Hotspots –  FuncKonal/pathway informaKon •  Deeper look at high priority groups of traits –  Gene set enrichment –  CorrelaKon with key clinical traits Correlated Traits UCLA Networks (c) Yandell 2013 34 MulK-­‐trait strategy (cont.) •  Use geneKcs to fine-­‐tune search –  Test Causal pairs (Elias talk, Mark on Tuesday) –  Screen out genes based on IBD, SNP effects •  IdenKfy small subset (5-­‐15) for networks –  Infer causal network (Elias talk) –  Use biological pathway informaKon (PPI, TF, …) –  Avoid hairballs (but hubs may be useful) •  ValidaKon tests of key drivers: new study … Correlated Traits UCLA Networks (c) Yandell 2013 35 how to use funcKonal informaKon? •  funcKonal grouping from prior studies –  may or may not indicate direcKon –  gene ontology (GO), KEGG –  knockout (KO) panels –  protein-­‐protein interacKon (PPI) database –  transcripKon factor (TF) database •  methods using only this informaKon •  priors for QTL-­‐driven causal networks –  more weight to local (cis) QTLs? Modules/Pathways SISG (c) 2012 Brian S Yandell 36 Thanks! •  Grant support –  NIH/NIDDK 58037, 66369 –  NIH/NIGMS 74244, 69430 –  NCI/ICBP U54-­‐CA149237 –  NIH/R01MH090948 •  Collaborators on papers and ideas –  Alan Aye & Mark Keller, Biochemistry –  Karl Broman, Aimee Broman, ChrisKna Kendziorski Modules/Pathways SISG (c) 2012 Brian S Yandell 37