This document discusses quantitative trait locus (QTL) mapping and the inclusive composite interval mapping (ICIM) method. It begins with an overview of quantitative traits, QTL mapping, and different mapping populations. It then describes problems with previous interval mapping methods and introduces the theoretical basis and methodology of ICIM, which can detect additive and interacting QTL while avoiding biased estimates. The document highlights several publications that have used ICIM in crops like rice, wheat, soybean, and maize. It concludes with an overview of the biparental population (BIP) functionality in the QTL IciMapping software, which implements six different QTL mapping methods including various interval mapping approaches and single marker analysis.

1. 1
Principle of QTL mapping and
Inclusive Composite Interval
Mapping (ICIM)
Jiankang Wang
CIMMYT China and CAAS
E-mail: wangjk@caas.net.cn or jkwang@cgiar.org
The 9th Workshop on QTL Mapping and Breeding Simulation
The University of Sydney, Cobbitty NSW, 7-9 March 2012

2. Outlines
 Quantitative traits and QTL mapping
 Inclusive composite interval
mapping (ICIM) for additive and
interacting QTL
 Selected publications using ICIM
 The BIP functionality in QTL
IciMapping
2

3. 1. Quantitative traits and
QTL mapping
3

4. Quantitative traits
 Continuous phenotypic variation
 Affected by many genes
 Affected by environment
 Epistasis
 Polygene (or multi-factorial )
hypothesis
 Classical quantitative genetics
4

5. What is QTL Mapping?
 The procedure to map individual genetic factors
with small effects on the quantitative traits, to
specific chromosomal segments in the genome
 The key questions in QTL mapping studies are:
 How many QTL are there?
 Where are they in the marker map?
 How large an influence does each of them
have on the trait of interest?
5

6. 6
Dataset of QTL mapping
 Mapping population
 Marker data of each individual in
the mapping population
 Linkage map
 Phenotypic data

7. 7
Example: 10 RIL of Rice
(linkage map of Chr. 5 )
Marker C263 R830 R3166 XNpb387 R569 R1553 C128 C1402 XNpb81 C246 R2953 C1447
Grain
width
(mm)
Position
(cM)
0.0 3.5 8.5 19.5 32.0 66.6 74.1 78.6 81.8 91.9 92.7 96.8
RIL1 0 0 0 0 0 0 0 0 0 0 0 0 2.33
RIL2 2 2 2 2 2 0 0 0 0 2 2 2 1.99
RIL3 0 2 2 2 2 2 2 2 2 2 2 2 2.24
RIL4 0 0 0 0 0 0 2 2 2 2 2 2 1.94
RIL5 0 0 0 0 0 2 2 0 0 0 0 0 2.76
RIL6 0 0 0 2 2 2 2 2 2 2 2 2 2.32
RIL7 0 0 0 0 0 0 0 0 0 0 0 0 2.32
RIL8 2 2 0 2 2 0 0 0 0 2 2 2 2.08
RIL9 0 0 0 0 2 2 0 0 0 0 0 0 2.24
RIL10 0 0 0 0 2 2 0 0 0 0 0 0 2.45

8. 8
Classification of mapping
populations
 Bi-parental mapping populations (linkage
mapping)
 Temporary population: F2 and BC
 Permanent population: RIL, DH, CSSL
 Secondary population
 Association mapping
 Natural populations: human and animals

9. 9
Overview on
QTL mapping methods
Single marker analysis (Sax 1923; Soller et al. 1976)
The single marker analysis identifies QTLs based on the difference
between the mean phenotypes for different marker groups, but cannot
separate the estimates of recombination fraction and QTL effect.
Interval mapping (IM) (Lander and Botstein 1989)
IM is based on maximum likelihood parameter estimation and provides
a likelihood ratio test for QTL position and effect. The major
disadvantage of IM is that the estimates of locations and effects of QTLs
may be biased when QTLs are linked.
Regression interval mapping (RIM)
(Haley and Knott 1992; Martinez and Curnow 1992 )
RIM was proposed to approximate maximum likelihood interval mapping
to save computation time at one or multiple genomic positions.

10. 10
Composite interval mapping (CIM) (Zeng 1994)
CIM combines IM with multiple marker regression analysis,
which controls the effects of QTLs on other intervals or
chromosomes onto the QTL that is being tested, and thus
increases the precision of QTL detection.
Multiple interval mapping (MIM) (Kao et al. 1999)
MIM is a state-of-the-art gene mapping procedure. But
implementation of the multiple-QTL model is difficult, since the
number of QTL defines the dimension of the model which is
also an unknown parameter of interest.
Bayesian model (Sillanpää and Corander 2002)
In any Bayesian model, a prior distribution has to be
considered. Based on the prior, Bayesian statistics derives the
posterior, and then conduct inference based on the posterior
distribution. However, Bayesian models have not been widely
used in practice, partially due to the complexity of
computation and the lack of user-friendly software.

11. 11
Principle of QTL mapping
 Three marker types at one marker locus
A. 很可能存
在QTL和标
记的连锁
性状平均数
mm MMMm
B. 不一定存
在QTL和标
记的连锁
性状平均数
mm MMMm

12. 12
Backcrosses (P1BC1 and P2BC1)
of P1: MMQQ and P2: mmqq
BC1 BC2
Genotype Frequency
Genotypic
value
Genotype Frequency
Genotypic
value
MMQQ )1(2
1
r− m+a MmQq )1(2
1
r− m+d
MMQq r2
1
m+d Mmqq r2
1
m-a
MmQQ r2
1
m+a mmQq r2
1
m+d
MmQq )1(2
1
r− m+d mmqq )1(2
1
r− m-a

13. 13
Principle of single marker
analysis (P1BC1 as example)
 Two marker types:
 Difference in phenotype between the two types
MMQqMMQQMM )1( µµµ rr +−=
rdarmdmramr +−+=+++−= )1()())(1(
MmQqMmQQMm )1( µµµ rr −+=
drramdmramr )1())(1()( −++=+−++=
))(21(MmMM dar −−=− µµ

14. 14
Interval mapping (IM)
(Lander and Botstein 1989)
 Linear model (j=1，2，…，n )
b* represent QTL effect， is the indicator
variable (0 or 1) for QTL genotype
 Likelihood profile
 Support interval: One-LOD interval
*
jx
jji exbby ++= **
0

15. 15
QTL genotypes under each marker
type in P1BC1 (double crossover not considered)
P1: P2:
F1: P1:
区间标记类型1 区间标记类型2 区间标记类型3 区间标记类型4
Mi Q Mi +1
Mi Q Mi +1
mi q mi +1
mi q mi +1
×
Mi Q Mi +1 Mi Q Mi +1
Mi Q Mi +1
×
Mi Q Mi +1 Mi Q Mi +1 Mi Q Mi +1 Mi Q Mi +1
Mi Q Mi +1 Mi Q mi +1 mi q mi q mi +1
Mi Q Mi +1
Mi q mi +1
mi q Mi +1
Mi Q Mi +1
mi Q Mi +1
mi q mi +1
Marker class I Marker class II Marker class III Marker class IV

16. 2. Inclusive Composite
Interval Mapping (ICIM)
16

17. 17
Problems with IM
 Assumption: No more than one QTL
per chromosome or linkage group
 “Ghost QTL” for linked QTL
 Large confidence interval
 Biased effect estimation
 Composite interval mapping (CIM)
(Zeng 1994)

18. 18
Problems with CIM
In the algorithm of CIM, both QTL effect at the
current testing position and regression coefficients
of the marker variables used to control genetic
background were estimated simultaneously in an
expectation and maximization (EM) algorithm.
Thus, this algorithm could not completely ensure
that the effect of QTL at current testing interval
was not absorbed by the background marker
variables and therefore may result in biased
estimation of the QTL effect.

19. 19
Theoretical basis of ICIM
∑∑ <=
+=
kj
kjjk
m
j
jj ggaagaG
1
1)|( ++= jjjjj xxgE ρλX
1111)|( ++++ +++= kjkjkjkjkjkjkjkjkj xxxxxxxxggE ρρλρρλλλX
i
kj
ikijjk
m
j
ijji exxbxbby +++= ∑∑ <
+
=
1
1
0

20. 20
Genomic scanning for
additive and interacting QTL
 Two-dimensional scanning (interval mapping)
 One-dimensional scanning (interval mapping)
∑+≠
−=∆
1,
ˆ
kkj
ijjii xbyy
∑∑
+≠
+≠++≠
−−=∆
1,
1,1,,1,
ˆˆ
kks
jjr
isirrs
kkjjr
irrii xxbxbyy

21. 0
10
20
30
40
11111111111222222222233333333334444444444
LODscore
Scanning posoition along the genome
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
11111111111222222222233333333334444444444
Effect
Scanning posoition along the genome
0
20
40
60
80
11111111111222222222233333333334444444444
LODscore
Scanning posoition along the genome
-4
-3
-2
-1
0
1
2
3
11111111111222222222233333333334444444444
Effect Scanning posoition along the genome
0
10
20
30
40
50
60
70
11111111111222222222233333333334444444444
LODscore
Scanning posoition along the genome
-1.5
-1
-0.5
0
0.5
1
1.5
11111111111222222222233333333334444444444
Effect
Scanning posoition along the genome 21
IM
CIM
ICIM

22. 22
Detecting
epistasis where
the interacting
QTL don’t have
significant
additive effects

23. 105
120
40
80
140
1, -
10, -
155
4, -
9, -
4, +
90
1, +
0
60
9, -
5, -
180
2, -
5, +
200
1, -
9, +
95
1, -
2, -
305
2, +
65
1, +
7, +
15
5, +
30 9, +
5, -
10, +
166
2, -
1 TL03BWW
2 TL03AIS
4 TL04BWW
5 TL04AIS
7 ZW03BIS
8 ZW03BSS
9 ZW04AWW
10 ZW04BSS
qFFLW1-1
qFFLW1-2
qFFLW1-3
qFFLW2
qFFLW3-1
qFFLW3-2
qFFLW4-1
qFFLW4-2qFFLW4-3qFFLW5-1
qFFLW5-2
qFFLW7-1
qFFLW7-2
qFFLW8-1
qFFLW8-2
qFFLW8-3
qFFLW10-1
qFFLW10-2
23
Digenic epistatic networks of FFLW in
maize using a RIL population

24. 3. Selected publications
using ICIM
24

25. In rice
 Crop Science (2008) 48: 1799-1806; Tiller
angle
 Hereditas (2009) 146: 67-73; Brown
planthopper resistance
 Mol. Breeding (2010) 25: 287-298; Heading
date
 Scientia Agricultura Sinica 2010,43(21):
4331-4340; Nitrogen efficiency
25

26. In wheat
Euphytica (2009) 165: 435-444; flour and noodle
color components and yellow pigment content
26

27. More in wheat
 Acta Agronomica Sinica (2011) 37 (2): 294-301; Coleoptile
Length and Radicle Length
 Crop & Pasture Science (2009) 60: 587-597; White salted
noodle quality
 Crop & Pasture Science (2011) 62: 625-638; Kernel morphology
traits
 Mol. Breeding (2010) 25: 615-622; Adult-plant resistance to
powdery mildew
 Theor. Appl. Genet. (2009) 119: 1349-1359; Adult-plant
resistance to stripe rust
 Mol. Breeding (2011) on line published; Grain protein content
and grain yield component
 Scientia Agricultura Sinica 2011,44(14):2857-2867; Grain yield
per plant and plant height
27

28. In soybean
 Breeding Science
(2008) 58: 355-359 ;
Salt tolerance
 ACTA
AGRONOMICA SINICA
2009, 35(12):
2139−2149; Protein
Related Traits
28

29. In Maize
 Theor. Appl. Genet. (2011) 123: 327-338;
Partial restoration of male fertility of C-type
cytoplasmic male sterility
 Plant Mol. Biol. Rep. (2011) on line published;
Nitrogen Use Efficiency
 HEREDITAS 32(6): 625-631; The area of
leaves
29

30. In melon
30
 Mol Breeding (2011) 27: 181-192; Powdery mildew

31. Publications using RSTEP-LRT
 In Rice:
Theor Appl Genet (2006) 112: 1258-1270; Grain length
Plant Cell Report (2009) 28: 247-256; Mature seed
culturability
Mol. Breeding (2010) 25: 287-298; Heading date
 Plant Cell Rep. (2009) 28: 247-256; Mature seed
culturability
 In Maize:
 Scientia Agricultura Sinica (2011),44(17):3508-3519;
Yield
31

32. 4. The BIP functionality in
QTL IciMapping
32

33. Six methods in BIP
 SMA: single marker analysis (Soller et al., 1976. Theor.
Appl. Genet. 47: 35-39)
 IM-ADD: the conventional simple interval mapping
(Lander and Botstein, 1989. Genetics 121: 185-199)
 ICIM-ADD: inclusive composite interval mapping of
additive (and dominant) QTL (Li et al., 2007. Genetics
175: 361-374. Zhang et al., 2008. Genetics 180: 1177-
1190)
 IM-EPI: interval mapping of digenic epistatic QTL
 ICIM-EPI: inclusive composite interval mapping of
digenic epistatic QTL (Li et al., 2008. Theor. Appl.
Genet. 116: 243-260)
 SGM: selective genotyping mapping (Lebowitz et al.,
1987. Theor. Appl. Genet. 73: 556–562)

34. Interface of the BIP functionality
Menu Bar Tool Bar
Project
Window
Message
Button
Task List
Button
Display
Window
Parameter
Setting Window

35. LOD profile of ICIM additive
mapping (ICIM-ADD)

36. Figures of interacting QTL from
ICIM epistatic mapping (ICIM-EPI)