UW ENCODE Scientific Presentation

yalechurlishAI and Robotics

Nov 7, 2013 (3 years and 9 months ago)

140 views

UW ENCODE Scientific
Presentation
ENCODE Consortium Call
20Feb
2009
20
 
Feb
., 
2009
John Stamatoyannopoulos, M.D.
Depts. of Genome Sciences and Medicine
University of Washington
Topics
2
1)   Digital DNaseIdata:  Signal, hotspots, DHSs
2)Properties:  Distribution, relation to gene expression
UW ENCODE digital DNaseI
3
In browser:
Cell types
ENm002
(1Mb)
K562
HepG2
HelaS3
GM06990
GM06990
SKnSH
Th1
More coming
soon 
(inclGM12878
(incl

GM12878
and HUVEC)
UW ENCODE digital DNaseI
4
Data types
Raw signal
Hotspots
&
 
FDR 0.5% Peaks
(=DHSs)
UW ENCODE digital DNaseI:    Hotspots vs. DHS Peaks
5
Hotspots= Genome‐wide
significant DNaseIcleavage
s
iteclusters@
zscore
>2
s
ite
 
clusters
?
@
?
zscore
>2
Peaks= Peaks within
FDR05%htt
FDR
 
0
.
5%
 
h
o
t
spo
t
s
UW ENCODE digital DNaseI:    Hotspots vs. DHS Peaks
6
Hotspots
FDR 0.5% Peaks
160,72489,752
209,337
136,005
76,558
88,245
235,102
131,301
131,440
123,89181,400
85,248
UW ENCODE digital DNaseI:    Verification/Reproducibility criteria
7
Data are highly
reproducible both
qualitatively
(positionsof
(positions
 
of
hotspots/DHSs)
and quantitatively
(magnitude of raw
il)
s
i
gna
l)
UW ENCODE digital DNaseI:    Verification/Reproducibility criteria
Verifieddata=RawsignalcorrelationcoefficientR>095acrossreference
8
Verified
 
data
?
=
?
Raw
?
signal
?
correlation
?
coefficient
?
R
?
>
?
0
.
95
 
across
?
reference
chromosome (Chr19)
Chr. 19
Fraction DHSs verified
2636
p
licate A
8
R
= 0.9806
Cell type
K562
9610
9631
A B
B A
Biological Re
p
DS#774
8
K562
.
9610
 .
9631
HepG2.9394 .9444
2578
0
0
Biological Replicate B
DS#7784
GM06990    .9373 .9415
Key advantage of using simple verification criterion based on raw data
:   Avoids issues of
algorithms, peak calling, thresholding, copy# artifacts, etc.
Topics
9
1)   Digital DNaseIdata:  Signal, hotspots, DHSs
2)Properties:  Distribution, relation to gene expression
Organization of cell‐specific regulatory DNA
10
ENm009
(1Mb)
Organization of cell‐specific regulatory DNA
11
promotersdistal
unusual
common
Marked cell type‐specificity
12
Cell type‐specific
Mixed lineage
Constitutive
Marked cell type‐specificity
13
Clustering of cell type‐specific DHSs into ‘regulons’
14
B‐lymphoblast
(GM06990)
Liver
(HepG2)
Th1
Erythroid
(K562)
Skin
(BJ Fibroblast)
HeLa
HeLa
Clustering of cell type‐specific DHSs into ‘regulons’
15
B‐lymphoblast
(GM06990)
(GM06990)
Liver
(HepG2)
Th1
Th1
Erythroid
(K562)
SkeletalMuscle
Skin
(BJ Fibroblast)
Skeletal
 
Muscle
Myeloid
Neuronal
(SKnSH + RA)
(HL60)
HeLa
Regulonsdefine archetypal cell and lineage‐specific genes
Exam
p
le:   Th1 re
g
ulons, GO classes of intersectin
g
 
g
enes:
16
p
g
gg
Exam
p
le:   He
p
atocellularre
g
ulons
Regulonsdefine archetypal cell and lineage‐specific genes
17
B-lymphoblast
(GM06990)
Liver
p
p
g
Liver
(HepG2)
Th1
Erythroid
(K562)
Skin
(BJFibroblast)
(BJ

Fibroblast)
Skeletal Muscle
Myeloid
Neuronal
(SKnSH + RA)
Myeloid
(HL60)
HeLa
Exam
p
le:   He
p
atocellularre
g
ulons, GO classes of intersectin
g
 
g
enes:
Regulonsdefine archetypal cell and lineage‐specific genes
18
p
p
g
gg
100kb
Regulonsand gene expression programming
19
B‐lymphoblast
Hepatic
Th1
Erythroid
ssion
B‐lymphoblast
(GM06990)
Fibroblast
Myoblast
Myeloid
HeLa
Expre
Hepatic
(HepG2)
Th1
Erythroid
Erythroid
(K562)
Fibroblast
(BJ)
Myoblast
Chromatin
Neuronal
(SKnSH+ RA)
Myeloid
(HL60)
HeLa
Regulonsand gene expression programming
HNF4a
20
B‐lymphoblast
Hepatic
Th1
Erythroid
ssion
B‐lymphoblast
(GM06990)
Fibroblast
Myoblast
Myeloid
HeLa
Expre
Neuronal
Hepatic
(HepG2)
Th1
Erythroid
Erythroid
(K562)
Fibroblast
(BJ)
Myoblast
Chromatin
Neuronal
(SKnSH+ RA)
Myeloid
(HL60)
HeLa
Regulonsand gene expression programming
Rl
dfill
ifid
21
R
egu
l
ons
d
e
fi
ne ce
ll
‐spec
ifi
c expresse
d
 genes
Acknowledgements
22
Digital DNaseImapping
Lab team:   Mike Dorschner, Peter Sabo, Molly Weaver, Jeff Goldy, Andrew 
Hdk
KitiLShi
V
TShfJ
Ni
H
ay
d
oc
k

K
r
i
s
ti

L
ee, 
Shi
nny 
V
ong, 
T
ony 
Sh
a
f
er, 
J
un 
N
er
i
Informatics team:  Bob Thurman, Scott Kuehn, Richard Sandstrom, Shane 
Ne
p
h
,
 Alex Re
y
nolds
,
 Richard Humbert
,
 Brendan Henr
y,
 Eric R
y
nes
,
 Eric Hau
g
en
p,y,,y,
y
,g
Fellows:   Xiangdong Fang, HaoWang, Sean Thomas, Brady Miller, Xia You
UW Genome Sciences, Medical Genetics & Immunology
Bill Noble, Evan Eichler, Chris Wilson, George Stamatoyannopoulos, Pat Navas