An introduction
Connectivity Map: the use for drug-repurposing
Speaker Wu-Lung Roger Yang 楊伍隆
Advisors Prof. Chi-Ying Huang ⿈黃奇英
Prof. Kun-Mao Chao 趙坤茂
Central Dogma
•DNA
•DNA
•RNA•Protein
•DNA
•RNA•Protein
DNA $$
RNA$
Protein$
transcription
translation
duplication
Extended central dogma
copy number variation SNP
Technology available
DNA
Sequencing, Next generation sequencing (NGS)
SNP array RNA
microarray NGS
Protein
protein array
mass spectrometry
DNA $$
RNA$
Protein$
References
Science 313(5795): 1929-1935.
Nature Reviews Cancer 7(1): 54-60.
Cited in Scopus: 750 (as May 4, 2013)
Cited in Scopus: 137 (as May 4, 2013)
Goals
Create a large drug- response profile
database.
Use microarray profiles
to query best matching
or mismatching drugs
Drug repurposing
http://2.bp.blogspot.com/_cwleTqQx7ks/TIkHQJSm-uI/AAAAAAAAABU/feudFmSg4MY/s1600/DNA+microarray.jpg
Microarray
Goals
Control Treatmen t
Cell Cell + Drug
Gene Control Treatme nt Fold Change
A 1 5 5/1
B 1 1 1/1
C 5 1 1/5
...
measurement of
mRNA
Drug 1 Drug 2 Drug 3 Drug ... Drug N
Gene 1 Gene 2 Gene 3 Gene 4 Gene 5 Gene 6 Gene ...
Gene M
Drug Similarity
Compare
Drug Repurposing
http://www.msd.com.hk/health_info/drug_education/e_ddp_introduction.html
Drug Development Process
Ashburn, Ted T, and Karl B Thor. 2004. “Drug repositioning: identifying and developing new uses for existing drugs.” Nature Reviews Drug Discovery
Nature Reviews Drug Discovery 12, 87-90 (February 2013) | doi:10.1038/nrd3946
2007: 69 drugs, but 16 are new drugs
average 15 years and US$800 million for a new drug to market
New drugs approved by FDA each year remain at 20~30 compounds.
New uses for old drugs
Nobel laureate James Black
Nature 448, 645-646 (2007)
"The most fruitful basis for the discovery of a new drug is to start with an old drug"
Chong, Curtis R, and David J Sullivan. 2007. “New uses for old drugs.” Nature 448(7154): 645–646.
New uses for old drugs
“The most fruitful basis for the discovery of a new drug is to start with an old drug”
-- Nobel laureate James Black
Drug repurposing examples
Viagra for erectile dysfunction Nelfinavir for cancer
Tamoxifen for bipolar disorder
Gleevec for rheumatoid arthiritis
Pentylenetetrazole for Down Syndrome Astemizole for malaria
Lipitor for alzhemimers
Lipitor for influenza mortality Metformin for cancer
CDD Community Group Meeting, SFO, Oct 1, 2009
Database
Drugs
1309 small molecules
680 drugs have
been conducted 4 times each.
6100 instances
7056 microarrays
0 175
350 525 700
49 43
162
680
226
87
13 17 3
1 2 3 4 5 6 7 8 9 Num of Instances
Drug Co unt
Drug dosage
Most drugs are tested between 1µM and 20µM
Co un t
Dose count
0 375 750 1125 1500
Number of doses
6 5 4 3 2 1
1225
10 72 1
0 1
Instance Count v.s. Number of Doses
Duration
Most drugs are conducted for 6 hours.
Primary effects, not second effects
0 1750 3500 5250 7000
6 Hours 12 Hours
4
6096
Cell Type
0 1000 2000 3000 4000
MCF7 PC3 HL60 ssMCF7 SKMEL5
17 18
1229 1741
3095
MCF7 Breast cancer PC3 Prostate Cancer
HL60 Leukemia
ssMCF7 Breast cancer
charcoal-stripped serum
SKMEL5 Skin Cancer
Fold Change
64% of 6100 instances have less than 2% of probes having 2-Fold change.
Percentage of probes with 2-Fold change
0 750 1500 2250 3000
<1%
1%~2%
2%~3%
3%~4%
4%~5%
5%~6%
6%~7%
7%~8%
8%~9%
>10%
128 197 80 70
219 132 381
1005 1967 2021
Instances
Each instance is defined as one drug treatment over a sample of controls.
e.g. cells treated with vorinostat versus untreated ones.
Instance = Drug type x Cell type x Duration x Dose
e.g. 10µM vorinostat on MCF7 for 6 hours
6100 instances over 1309 drugs
Cel l T ype
Drug
Do sag e
2008 price: $250 US
Lamb, J et al. 2006. “The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease.” Science 313(5795):
Lamb, Justin. 2007. “The Connectivity Map: a new tool for biomedical research.” Nature Reviews Cancer 7(1): 54–60.
http://www.broadinstitute.org/cmap/
Drug-Drug similarity
Input: 13 HDAC gene signature
Lamb, J et al. 2006. “The Connectivity Map: using gene-expression signatures to connect smallDrug-disease
Sirolimus reverses gluco-corticoid resistance in acute lymphoblastic leukemia.
Lamb, J et al. 2006. “The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease.” Science 313(5795): 1929–1935.
Dexamethasone only
Dexamethasone + Sirolimus
Sample Calculation
Kolmogorov-Smirnov statistics
The score captures the
maximum derivation from the random distribution.
0 0.25 0.5 0.75 1
10 2 3 4 5
0.1
0.5
0.8
0.9
0.2
0.4
0.6
0.8
1
213731_s_at 208793_x_at 208711_s_at 208623_s_at
203725_at
200965_s_at 201540_at 215677_s_at 205472_s_at
204802_at
Sample calculation
-‐0.608
Raw score calculation
Score
• t: number of tags
• n: total number of tags (22283)
• score = a if a >b
• score = -b if b>a
b =13542/22283 - 0/5 = 0.608 a = 4/5 - 2801/22283 = 0.674
-‐0.608
Drug ranking
Enrichment
• t: number of instances
• n: total number of instances (6100)
5171/6100 - 0/1 = 0.848
Permutation
p-value is estimated through permutation analysis by
constructing empirical null
distribution.
http://rlv.zcache.com/my_p_value_is_smaller_than_your_p_value_tshirt-rdb18d031a5774ce18f1a42edfe212916_804gs_512.jpg