1

Term ReportTerm ReportTerm ReportTerm Report

2

Next week

• Protein-protein interaction– Yeast two-hybrid (Nature 2000)– TAP-tag + MS (Nature 2002)– FLAG-tag + MS (Nature 2002)– Pooling strategy (Science 1999)

3

Now what ?

• 576 strains, 20 growth conditions• Data, data, data….• Look for extremes ?• Cluster

– Group together similar patterns– Mathematical description

• Co-expression• Standard correlation coefficient

– Graphical representation• Original experimental observation• Color, dark and light• Visualize and understand the relationships intuitively

1,2,3,4,…….……,8,9,10

4

Data analysis

• Transformants clustering

• Graphical representation

Growth conditions

Transformants

http://bioinfo.mbb.yale.edu/genome/phenotypes/

5

Data clusters• 20 Growth conditions

6

Double cluster• Horizontal cluster: transformants• Vertical cluster: growth conditions• Identify assays for functionally related proteins

7

Growth condition cluster

• Clustering growth conditions that result in similar phenotypes• More effective screening functionally related proteins

DNA metabolism

Cell wall biogenesis and maintenance

8

Discovery Questions

• What advantage is there to clustering the phenotypes in this manner?

• Some of the genes identified in this analysis had no known function. How can clustering these data help us predict possible functions?

9

PCR-based Gene Deletion • Deletion strategy

http://www-sequence.stanford.edu/group/yeast_deletion_project/PCR_strategy.html

10

Primers• 8~10 different primers/ORF• High-throughput primer synthesis

– Primer-picking scripts– Input:

• ORF data + UPTAG list

– Output:• Primer sequences• Automated Multiplex Oligonucleotide Synthesizer

• Homology to ORF upstream

• Common tag priming site (U1)

• UPTAG (20 bases)

• Common tag priming site (U2) homologous to 5’ to the Kan gene

11

Functional assays• 6925 deletion strains of yeast constructed…• 2026 ORF (1/3 of the genome)• Screen for genes essential for viability

– Spores from heterozygous strains on YPD media at 30oC• 356 haploid deletants could not be recovered• 1620 ORFs not essential for viability in yeast

– Construct one additional homozygous and two haploid deletants

Essential gene: tall bar Nonessential: short bar

12

Competitive growth assays• How to characterize the genes nonessential for viability?• Pooled functional assay

– 558 homozygous deletion strains were pooled– Grow in rich and minimal media for ~60 generations– Remove aliquots from the two pools at various time points– Tags were amplified and hybridized to DNA array

Red: grown for 0 hr Green: grown for 6 hr

Normal growth (expression)

Grow slow (reduced expression)Grow fast (enhanced expression)

13

Correlation of growth rate

• Where is the wild type ?

In rich medium

r = 0.97

In minimal medium

r = 0.94

14

Normalized growth rate

In rich medium

• Hybridization intensity = growth rate– Normal growth = 1– Grown fast (abundant) > 1– Grown slow (fewer) < 1

p. 170, Q22:

Predict what might happen if only the slowest growing strains were incubated together.

15

Comparison

• mTn method• Pros:• Cons:

• Bar code method• Pros:• Cons:

16

Completed genome

• Unicellular eukaryotes– Budding yeast, Saccharomyces

cerevisiae

• Multicellular eukaryotes– Nnematode, Caenorhabditis elegans– Fruit fly, Drosophila melanogaster

17

RNAi

• RNA interference by Andy Fire, 1998• RNAi transiently inhibits the activity of

a target gene with a dsRNA

18

RNAi and C. elegans

• C. elegans eats E. coli expressing specific dsRNA

• Observe phenotypes of adult and embryo development

E. Coli with Gene A

B

C

19

Functional Distribution

• Genes on chromosome I of C. elegans

Fraser et al., 2000 Nature 408, 325-330.

20

Comparisons• The British group

– Chromosome I– Bacterial expressed dsRNA– By feeding– Viability, and observable phenotypes

• The German group– Chromosome III (cell division process)– PCR amplified, in vitro transcription ssRNA, annealed to

generate dsRNA– By microinjection

• RNAi strategy– Pros– Cons