Transcript Complex trait analysis, develop

The GeneNetwork and WebQTL : PART 2
link to www.genenetwork.org
Part 1. How to study
expression variation
and genetic
correlation (slides 2–
17)
Part 2. Discovering
upstream modulators
(slides 18–29)
RNA
How to make recombinant inbred strains (RI)
female
C57BL/6J (B)
fully
inbred
male
DBA/2J (D)
BXD
chromosome pair
isogenic
F1
heterogeneous
F2
Inbred
Isogenic
siblings
20 generations
Recombined
chromosomes
are needed for
mapping
BXD RI
Strain set
BXD1
brother-sister
matings
BXD2
+…+
BXD80
UPSTREAM
modulators
D B
trans QTL
High
4 units
aaaa
amount of
transcript
Low
D2 strain
2 units
aa
B6 strain
>>>>PROMOTER--ATG-Exon1-Intron1-Exon2-Intron2 - etc-3'UTR >>>>>
D B
cis QTL
D and B may be SNP-like variants in the promoter
itself (cis QTL) or in upstream genes (trans QTLs).
Discovering upstream modulatory loci
WebQTL searches for upstream controllers
App maps on Chr 16 (blue
arrow points to the orange
triangle) but the best locus
is on Chr 7.
Genetic versus Physical maps for App expression
The difference between genetic and physical scale is analogous to measuring the separation
between New York and Boston in either travel hours or kilometers.
Physical map for distal chromosome 7
Distal Chr 7 from ~120 and 132 Mb may modulate App
Evaluating candidate genes
Right position
and high
correlation
= better
candidates
Physical maps are zoomable
Evaluating Ctbp2 as a candidate QTL for App
This is the App QTL
in the INIA data set.
This is the Ctbp2
cis QTL, but is
detected only in
the Rosen
striatum data
set.
Evaluating Ctbp2 using other resources
Summary of Part 2
1. Covered the basics of QTL analysis and mapping.
2. Reviewed difference between genetic and physical maps.
3. Discussed interpreting features of QTL maps including the LRS
function, the additive effect function, the bootstrap bars, and the
permutation thresholds.
4. Illustrated techniques to generate a list of positional candidates.
5. Discussed some factors used to evaluate candidate genes.
What does a QTL signify? A good QTL is a claim that a particular
chromosomal region contains a causal source of variation in the
phenotype. The importance of this hypothesis depends on the quality
and relevance of the phenotype and the statistical strength of the QTL.
As usual, test and be skeptical.
Test Questions
1. Evaluate candidates for the Chr 3 App QTL.
2. Do App and Ctbp2 expression share any
other QTLs beside that on Chr 7?
3. Can you exploit literature mining tools to
find a strong relationship between App and
Ctbp2?
4. Why might the cis QTL for Ctbp2 expression
only be detected in the striatum data set?
Contact for comments and improvements:
[email protected]
[email protected]
The App findings reviewed in this presentation are part of an
ongoing study by R. Williams. R. Homayouni, and R. Clark
(July 15, 2005)