Transcript Undergraduates Learning Genomics Through Research

Undergraduates Learning Genomics
Through Research
A. Malcolm Campbell
University of Wisconsin - Madison
May 15, 2008
Seven Year Collaboration; Three Countries
www.bio.davidson.edu/GCAT
GCAT Makes DNA Chips Affordable
Steady Growth Over Time
10,000+ Undergraduates and Counting
Distribution of GCAT Members
GCAT Publication of Outcomes
Basic Research Publications
2008: 4 peer-reviewed publications
2007: 2 peer-reviewed publications
2006: 1 peer-reviewed publication
Student Learning Outcomes
Question
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Topic
Microarray experimental error–dye bias
Microarray experimental error–gradient
Microarray negative controls
Microarray experimental design
Gene expression ratios using a graph
Gene expression–probability
Gene expression–gene clusters
Gene expression–regulatory cascade
Gene expression–gene circuit graphs
Interpreting microarray results
Diagnosis with microarrays
* indicates p < 0.05; N = 409
Increase (%)
+ 36.2*
+ 10.5*
+ 10.3*
+ 38.2*
+ 5.8*
+ 0.2
+ 22.3*
+ 14.9*
+ 11.8*
+ 19.0*
+ 12.5*
Increased Student Interest in Research
. Area
Genomics
Biology
Research
1 = decreased a lot
7 = increased a lot
N = 409
Mean
5.5
5.5
5.4
SD .
1.1
1.1
1.2
Student Satisfaction with Methods
Activity
Practicing data analysis before my own data
Isolating RNA or genomic DNA to produce probe
Producing the fluorescently labeled probe
Hybridizing the probe with the spotted DNA
Designing my own experiment
Analyzing data from public domain source
Reading papers that used DNA microarrays
1 = not effective at all
7 = highly effective
Mean
5.25
5.32
% >4
93.6
94.1
% >5
67.1
70.0
N
313
323
5.22
5.20
5.13
5.22
5.06
94.4
92.8
87.3
94.7
88.9
68.9
70.1
64.3
65.8
62.4
306
334
244
325
343
Course Grade Assessment
Assessment method % Using method
Test
42.2
Term paper/lab report
51.1
Poster presentation
33.3
Oral presentation
26.6
Manuscript for publication
8.8
Course evaluation
33.3
Informal feedback
62.2
Other
24.4
Faculty Appreciate GCAT Resources
Mean
Access to microarray technology without GCAT1.5
0.75
Online GCAT protocols useful
4.4
The GCAT -Listserv helpful
4.2
SD
GCAT network significant factor
4.2
0.79
Positive experience using GCAT
I would use GCAT again in the future
4.6
4.7
0.60
0.63
1 = strongly disagree
5 = strongly agree
0.69
1.0
Faculty Development
“You have awakened parts of my brain that have
been dormant since my last stats course. The
only reason I have gone over the manual so
carefully is that this is my first time teaching
microarrays, or even using them, for that matter.
GCAT has been remarkably helpful to me. In fact
I don’t think I would have undertaken this new
module in my lab course without the tools GCAT
makes available.”
Introduced Microarrays Early
Ben Kittinger ‘05
Wet-lab microarray simulation kit - fast, cheap, works every time.
GCAT Develops Commercial Product
Enable Students to Practice
www.bio.davidson.edu/projects/GCAT/Spot_synthesizer/Spot_synthesizer.html
Open Source and Free Software
www.bio.davidson.edu/MAGIC
What Else Can Chips Do?
Jackie Ryan ‘05
Comparative Genome Hybridizations
Synthetic Biology
What is Synthetic Biology?
BioBrick Registry of Standard Parts
http://parts.mit.edu/registry/index.php/Main_Page
What is iGEM?
Peking University
Imperial College
iGEM Team Mosaic of Institutions
SYNTHETIC BIOLOGY
iGEM 2006
Davidson College
Malcolm Campbell (bio.)
Laurie Heyer (math)
Karmella Haynes (HHMI)
Lance Harden
Sabriya Rosemond (HU)
Samantha Simpson
Erin Zwack
Missouri Western State U.
Todd Eckdahl (bio.)
Jeff Poet (math)
Marian Broderick
Adam Brown
Trevor Butner
Lane Heard (HS student)
Eric Jessen
Kelley Malloy
Brad Ogden
Advantages of Bacterial Computation
Software
Hardware
Computation
Computation
Computation
Advantages of Bacterial Computation
Software
Hardware
Computation
$
Computation
¢
Computation
Burnt Pancake Problem
1
2
3
4
Burnt Pancake Problem
Look familiar?
Advantages of Bacterial Computation
• Non-Polynomial (NP)
# of Processors
• No Efficient Algorithms
Cell Division
Flipping DNA with Hin/hixC
Flipping DNA with Hin/hixC
Flipping DNA with Hin/hixC
How to Make Flippable DNA Pancakes
All on 1 Plasmid: Two pancakes (Amp vector) + Hin
RBS
pLac
Hin LVA
T
T
RBS
hixC
Tet
pBad hixC
pancake 1
hixC
pancake 2
Hin Flips DNA of Different Sizes
Hin Flips Individual Segments
-2
1
No Equilibrium 11 hrs Post-transformation
Hin Flips Paired Segments
mRFP off
1
-2
double-pancake flip
mRFP on
-1
white light
2
u.v.
Modeling to Understand Flipping
(-2,-1)
(-2,1)
(1,2)
(-1,2)
(1,-2)
(-1,-2)
(2,-1)
(2,1)
( 1, 2)
(-2, -1)
( 1, -2)
(-1, 2)
(-2, 1)
( 2, -1)
(-1, -2)
( 2, 1)
Modeling to Understand Flipping
(-2,-1)
(-2,1)
(1,2)
(-1,2)
(1,-2)
(-1,-2)
(2,-1)
(2,1)
1 flip: 0% solved
( 1, 2)
(-2, -1)
( 1, -2)
(-1, 2)
(-2, 1)
( 2, -1)
(-1, -2)
( 2, 1)
Modeling to Understand Flipping
(-2,-1)
(-2,1)
(1,2)
(-1,2)
(1,-2)
(-1,-2)
(2,-1)
(2,1)
2 flips: 2/9 (22.2%)
solved
( 1, 2)
(-2, -1)
( 1, -2)
(-1, 2)
(-2, 1)
( 2, -1)
(-1, -2)
( 2, 1)
Consequences of DNA Flipping Devices
-1,2
-2,-1
in 2 flips!
PRACTICAL
Proof-of-concept for bacterial
computers
Data storage
n units gives 2n(n!) combinations
BASIC BIOLOGY RESEARCH
Improved transgenes in vivo
gene
Evolutionary insights
Success at iGEM 2006
Living Hardware to Solve
the Hamiltonian Path Problem, 2007
Students:
Oyinade Adefuye,
Will DeLoache,
Jim Dickson,
Andrew Martens,
Amber Shoecraft, and
Mike Waters; Jordan
Baumgardner, Tom
Crowley, Lane Heard,
Nick Morton, Michelle
Ritter, Jessica Treece,
Matt Unzicker,
Amanda Valencia
Faculty: Malcolm Campbell, Todd Eckdahl, Karmella
Haynes, Laurie Heyer, Jeff Poet
The Hamiltonian Path Problem
1
4
3
2
5
The Hamiltonian Path Problem
1
4
3
2
5
Hin/hixC to
to Solve
the HPP
Using Using
Hin/hixC
Solve
the HPP
1
4
3
2
5
1 3
4 5
4 3
3 2 1 4 2 4
3 5
4 1
Hin/hixC to
to Solve
the HPP
Using Using
Hin/hixC
Solve
the HPP
1
4
3
2
5
1 3
4 5
4 3
3 2 1 4 2 4
hixC Sites
3 5
4 1
Hin/hixC to
to Solve
the HPP
Using Using
Hin/hixC
Solve
the HPP
1
4
3
2
5
Hin/hixC to
to Solve
the HPP
Using Using
Hin/hixC
Solve
the HPP
1
4
3
2
5
Using
Hin/hixC
Solvethe
the
HPP
Using
Hin/hixC to
to Solve
HPP
1
4
3
2
5
Using Hin/hixC to Solve the HPP
1
4
3
2
5
Solved Hamiltonian Path
How to Split a Gene
RBS
Detectable
Phenotype
Reporter
Promoter
RBS
Promoter
Repo-
rter
hixC
?
Detectable
Phenotype
Gene Splitter Website
http://gcat.davidson.edu/iGEM07/genesplitter.html
Input
Output
1. Gene Sequence
(cut and paste)
1. Generates 4 Primers
(optimized for Tm).
2. Where do you want
your hixC site?
2. Biobrick ends are
added to primers.
3. Pick an extra base to
avoid a frameshift.
3. Frameshift is
eliminated.
Gene-Splitter Output
Note: Oligos are
optimized for Tm.
Predicting Outcomes of
Bacterial Computation
Probability of HPP Solution
Starting Arrangements
4 Nodes & 3 Edges
Number of Flips
How Many Plasmids Do We Need?
Probability of at least k solutions on m plasmids for a 14-edge graph
k=1
5
10
20
m = 10,000,000
.0697
0
0
0
50,000,000
.3032
.00004
0
0
100,000,000
.5145
.0009
0
0
200,000,000
.7643
.0161
.000003
0
500,000,000
.973
.2961
.0041
0
1,000,000,000
.9992
.8466
.1932
.00007
k = actual number of occurrences
λ = expected number of occurrences
λ = m plasmids * # solved permutations of edges ÷ # permutations of edges
Cumulative Poisson Distribution:
e   x
P(# of solutions ≥ k) = 1 
x!
x0
k1
False Positives
Extra Edge
1
4
3
2
5
False Positives
PCR Fragment Length
1
4
3
2
5
PCR Fragment Length
Detection of True Positives
100000000.00
Total
Total##ofofPositives
Positives
10000000.00
1000000.00
100000.00
10000.00
1000.00
100.00
1
Total # of Positives
# of True Positives ÷
10.00
1.00
4/6
6/9
7/12
7/14
of Nodes / # of Edges
## of
Nodes / # of Edges
0.75
0.5
0.25
0
4/6
6/9
7/12
# of Nodes / # of Edges
7/14
How to Build a Bacterial Computer
Choosing Graphs
C
A
A
B
B
Graph 1
Graph 2
D
Splitting Reporter Genes
Green Fluorescent Protein
Red Fluorescent Protein
Splitting Reporter Genes
GFP Split by hixC
RFP Split by hixC
HPP Constructs
Graph 0 Construct:
A
AB
B
Graph 1 Constructs:
Graph 0
ABC
C
ACB
A
B
Graph 1
BAC
Graph 2 Construct:
DBA
A
B
Graph 2
D
Coupled Hin & HPP Graph
Hin +
Unflipped
HPP
Transformation
PCR to
Remove Hin
Ligate &
Transform
Flipping Detected by Phenotype
ABC
(Yellow)
ACB
(Red)
BAC
(None)
Flipping Detected by Phenotype
ABC
(Yellow)
ACB
(Red)
BAC
(None)
Hin-Mediated
Flipping
ABC Flipping
Yellow
Hin
Yellow, Green, Red, None
ACB Flipping
Red
Hin
Yellow, Green, Red, None
BAC Flipping
None
Hin
Yellow, Green, Red, None
Flipping Detected by PCR
ABC
ACB
BAC
BAC
ABC
ACB
Unflipped Flipped
Flipping Detected by PCR
ABC
ACB
BAC
BAC
ABC
ACB
Unflipped Flipped
Flipping Detected by Sequencing
BAC
RFP1
hixC
GFP2
Flipping Detected by Sequencing
BAC
RFP1
Flipped-BAC
RFP1
hixC
GFP2
Hin
hixC
RFP2
Conclusions
• Modeling revealed feasibility of our approach
• GFP and RFP successfully split using hixC
• Added 69 parts to the Registry
• HPP problems given to bacteria
• Flipping shown by fluorescence, PCR, and sequence
• Bacterial computers are working on the HPP and
may have solved it
Living Hardware to Solve the
Hamiltonian Path Problem
Acknowledgements: Thanks to The Duke Endowment, HHMI, NSF DMS 0733955,
Genome Consortium for Active Teaching, Davidson College James G. Martin
Genomics Program, Missouri Western SGA, Foundation, and Summer Research
Institute, and Karen Acker (DC ’07). Oyinade Adefuye is from North Carolina Central
University and Amber Shoecraft is from Johnson C. Smith University.
What is the Focus?
Thanks to my life-long collaborators
Extra Slides
Enter: Flapjack & The Hotcakes
Erin Zwack (Jr. Bio); Lance Harden (Soph. Math); Sabriya Rosemond (Jr. Bio)
Enter: Flapjack & The Hotcakes
Erin Zwack (Jr. Bio); Lance Harden (Soph. Math); Sabriya Rosemond (Jr. Bio)
Wooly Mammoths of Missouri Western
Genome Sequencing
Sarah Elgin at Washington University
Genome Education Partnership
Students finish and annotate genome sequences
Support staff online
Free workshops in St. Louis
Growing number of schools participating
Tuajuanda Jordan at HHMI
Phage Genome Initiative
Science Education Alliance
Students isolate phage
Students purify phage DNA; Sequenced at JGI
Students annotate and compare geneomes
National experiment to examine phage variation
Free workshop and reagents
Cheryl Kerfeld at Joint Genome Institute
Undergraduate Genomics Research Initiative
> 1000 prokaryote genomes sequenced
Students annotate genome
Data posted online
Workshop for training of faculty
Wide range of species
Burnt Pancake Problem
Design of controlled flipping
RBS-mRFP hix pLac hix
(reverse)
RBS-tetA(C)
hix
Can we build a biological computer?
The burnt pancake problem can be modeled as DNA
(-2, 4, -1, 3)
(1, 2, 3, 4)
DNA Computer Movie >>