t'Hoen

neighgreasycornerBiotechnology

Dec 14, 2012 (8 years and 10 months ago)

368 views

Deep sequencing
-
based expression
analysis shows major advances in
robustness, resolution and inter
-
lab
portability over five microarray
platforms

Peter A. C. ‘t
Hoen
, Johan T. den
Dunnen
,
et al.

Jared Taylor


Introduction


Methods


Results


Future

Feb 2012

Jared Taylor

2


Introduction


Methods


Results


Future

Feb 2012

Jared Taylor

3

Introduction


Compare hippocampal expression profiles



Gene expression microarrays



Illumina

Gene Expression Analysis



Bayesian modeling

Feb 2012

Jared Taylor

4

Goals


Define hippocampal expression



Reproducible
results across laboratories



Compare microarrays to
DGE/RNA
-
Seq



Prove
effectivity

of a new technology


Feb 2012

Jared Taylor

5

History


Gene Expression Microarrays



Serial Analysis of Gene Expression (SAGE)



δ
C
-
doublecortin
-
like kinase (DCLK)
-
short



Feb 2012

Jared Taylor

6


Introduction


Methods


Results


Future

Feb 2012

Jared Taylor

7

Animal Prep


C57/BL6j mice control



DCLK
-
short derived
from control



Hippocampi dissection


Feb 2012

Jared Taylor

8

RNA Extraction


Isolated using
TRIzol



Purified using
RNeasy



RNA quality assessed

Feb 2012

Jared Taylor

9

Sequenced


Digital Gene Expression (DGE) Tag Profile



Placed on an
Illumina

1G
flowcell



Illumina

Whole Genome Sequencer

Feb 2012

Jared Taylor

10

Sequence Tag Prep

Feb 2012

Jared Taylor

11

Data


Sequences tagged and counted



Gene
Expresssion

Omnibus


GSE10782



Tags annotated, only perfect matches kept

Feb 2012

Jared Taylor

12

Statistical Analysis


Compared to Microarrays


GSE8349



Bayesian model for biological variability



Feb 2012

Jared Taylor

13

qPCR

Assay


RNA samples identical to DGE RNA



SYBR
-
Green detection



Each
cDNA

sample ran four times

Feb 2012

Jared Taylor

14

Biological Pathway Analysis


Gene Ontology
-
defined pathway test


Bioinformatics initiative


Define genes and gene attributes across species

Feb 2012

Jared Taylor

15


Introduction


Methods


Results


Future

Feb 2012

Jared Taylor

16

Sequencing Statistics


2.4
±

1.5 x 10
6

sequences per sample



~2.0 x 10
5

unique tag sequences



70% Canonical, 20% of which are unique


Implies high amount of RNA

Feb 2012

Jared Taylor

17

Annotation Grouping

Feb 2012

Jared Taylor

18

Reproducibility


Pooled RNAs were analyzed at alternate lab



Pearson correlation coefficients used


-
1 to 1, 1 meaning the relationship is perfect



>.99 in same laboratory


.98 and .96 in other lab for wild
-
type and
transgenic

Feb 2012

Jared Taylor

19

Dynamic Range


Ckb

5.5 x 103 transcripts per million



Lowest 2 TPM, ~.3 per cell



Hippocampus had most unique tags



Feb 2012

Jared Taylor

20

Alternative
Polyadenylation


DGE can discern
polyadenylation

differences



47% of detected transcripts have >1 tag



Actual
polyadenylation

may be >47%



Feb 2012

Jared Taylor

21

Antisense Transcription


Evidence for 51% of bidirectional transcription



Antisense genes expressed



11% expressed greater than sense transcript



D
ecreased likelihood that antisense is an RT
artifact

Feb 2012

Jared Taylor

22

Differential Expression


Pooling samples proven to have negative data
impact



Bayesian Modeling found to better portray
expression profiles



1559 genes up expressed, 1620 down


8.5% false discovery rate

Feb 2012

Jared Taylor

23

Pooling Increases Error

Feb 2012

Jared Taylor

24

Bayesian Volcano Plot

Feb 2012

Jared Taylor

25

Biological Implication

Feb 2012

Jared Taylor

26

Differential Expression

Feb 2012

Jared Taylor

27


Introduction


Methods


Results


Discussion


Future

Feb 2012

Jared Taylor

28

DGE
vs

SAGE

Digital Gene Expression


3 Days



Unlimited* Tags



Differential Genes = 3179



Detection = 0.8 TPM


Serial Analysis of Gene Expression


1 Year



Max 100,000 Tags



Differential Genes = 200



Detection = 91 TPM


Feb 2012

Jared Taylor

29

DGE
vs

Microarrays

Feb 2012

Jared Taylor

30

DGE
vs

Microarrays

Feb 2012

Jared Taylor

31

DGE
vs

qPCR


62 genes assayed


43 showed concordant change


Only 5 significant on both technologies

Feb 2012

Jared Taylor

32

Conclusions


RNAseq

is the future


DGE currently more affordable



Low abundance transcript changes
undetectable on older technologies



DGE has less bias than SAGE and microarrays

Feb 2012

Jared Taylor

33


Introduction


Methods


Results


Future

Feb 2012

Jared Taylor

34

Future Research


Transcriptome

Analysis



Gene (Expression) Profiling



Genetic Engineering



Feb 2012

Jared Taylor

35

Further Reading


Power of Deep Sequencing and Agilent Microarray for Gene Expression
Profiling Study.
Molecular Biotechnology
, 2010: 45, 2, 101
-
110.



Applications of next generation sequencing in molecular ecology of non
-
model
organisms.

Heredity,

2011
:

107,

1

15.



Analysis of HIV
-
1 Expression Level and Sense of Transcription by High
-
Throughput Sequencing of the Infected
Cell.

Journal of Virology
,

July
2011:

85, 13,

6205
-
6211.



Programming cells by multiplex genome engineering and accelerated
evolution.
Nature
,
13 August 2009: 460, 894
-
898.

Feb 2012

Jared Taylor

36


Introduction


Methods


Results


Future

Feb 2012

Jared Taylor

37