Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
The New Genomics
Search
Matt Wood
October 02, 2012
Science
3
13k
The New Genomics
The value of reproducing, reusing and remixing scientific research. Slides from Strata London.
Matt Wood
October 02, 2012
Tweet
Share
More Decks by Matt Wood
See All by Matt Wood
Field Notes from Expeditions in the Cloud
mza
2
260
A Platform for Big Data
mza
6
640
The Data Lifecycle
mza
5
390
Provision Throughput Like a Boss
mza
0
340
Impact of Cloud Computing: Life Sciences
mza
2
740
Latency's Worst Nightmare: Performance Tuning Tips and Tricks
mza
4
990
Under the Covers of DynamoDB
mza
4
830
From Analytics to Intelligence: Amazon Redshift
mza
9
890
Scaling Science
mza
3
390
Other Decks in Science
See All in Science
OptimizationNight~機械学習と数理最適化の融合~
hidenari
0
350
Machine Learning for Materials (Lecture 9)
aronwalsh
0
130
統計的因果探索の方法
sshimizu2006
1
920
拡散モデルの概要 −§1. 拡散モデルで使われる確率微分⽅程式について−
nearme_tech
0
120
脳とAIは似ているか ― NeuroAI の挑戦
ykamit
9
7k
HAS Dark Site Orientation
astronomyhouston
0
5k
最新のAI技術を使った材料シミュレーションで材料研究現場に変革を
matlantis
0
550
B-Cubed: Leveraging analysis-ready biodiversity datasets and cloud computing for timely and actionable biodiversity monitoring
peterdesmet
0
180
ベクトル型スーパーコンピュータ「AOBA-S」の性能評価
keichi
0
310
BigQueryで参加するレコメンドコンペ / bq-recommend-competition-kaggle-meetup-tokyo-2023
shimacos
1
1.4k
Machine Learning for Materials (Lecture 8)
aronwalsh
0
340
Spark_Task_Optimization_Journey_How_I_Increased_10x_Speed_by_Performance_Tuning
tlyu0419
0
210
Featured
See All Featured
CSS Pre-Processors: Stylus, Less & Sass
bermonpainter
352
28k
We Have a Design System, Now What?
morganepeng
43
6.8k
A designer walks into a library…
pauljervisheath
201
23k
How to name files
jennybc
65
94k
Ruby is Unlike a Banana
tanoku
96
10k
Optimising Largest Contentful Paint
csswizardry
13
2.5k
The Cult of Friendly URLs
andyhume
74
5.7k
XXLCSS - How to scale CSS and keep your sanity
sugarenia
242
1.2M
Designing for Performance
lara
603
67k
YesSQL, Process and Tooling at Scale
rocio
165
13k
Embracing the Ebb and Flow
colly
80
4.2k
The Brand Is Dead. Long Live the Brand.
mthomps
49
32k
Transcript
The New Genomics
[email protected]
Dr. Matt Wood
Hello
Hello
Data
DNA
Chromosome 11 : ACTN3 : rs1815739
Chromosome X : rs6625163
Chromosome 19 : FUT2 : rs601338
+0.25 Chromosome 15 : rs2472297
Chromosome 2 : rs10427255
TYPE II Chromosome 10 : rs7903146
Chromosome 1 : rs4481887
I know this, because...
None
A T C G G T C C A G
G
A T C G G T C C A G
G A G C C A G G U C C Transcription
A T C G G T C C A G
G A G C C A G G U C C Translation Ser Glu Val Transcription
None
None
Chromosome 11 : ACTN3 : rs1815739
Chromosome X : rs6625163
Chromosome 19 : FUT2 : rs601338
+0.25 Chromosome 15 : rs2472297
Chromosome 2 : rs10427255
TYPE II Chromosome 10 : rs7903146
Chromosome 1 : rs4481887
I know all that, because...
Human Genome Project
40 species ensembl.org
Compare species
Biological importance
Step change
Less time. Lower cost.
None
None
Compare individuals
None
Data generation costs are falling (pretty much everywhere)
Sequencing challenge X
Amazona vittata
Analytics challenge
Lots of data, Lots of uses, Lots of users, Lots
of locations
Cost
Analytics challenge X
Accessibility challenge
The New Genomics
Graceful. Beautiful.
Impossible to re-create
Snowflake Science
Reproducibility
Reproducibility scales science
Reproduce. Reuse. Remix.
Value++
None
How do we get from here to there? 5PRINCIPLES REPRODUCIBILITY
OF
1. Use the gravity of data 5 PRINCIPLES REPRODUCIBILITY OF
Increasingly large data collections
1000 Genomes Project: 200Tb
Challenging to obtain and manage
Expensive to experiment
Large barrier to reproducibility
Data size will increase
Data integration will increase
Move data to the users
Move data to the users X
Move tools to the data
Place data where it can consumed by tools
Place tools where they can access data
None
None
None
Canonical source
None
More data, more users, more uses, more locations
Cost and complexity
Cost and complexity kill reproducibility
Utility computing
Availability
Intel Xeon E5 NVIDIA Tesla GPUs
90 - 120k IOPS on SSDs
Pay-as-you-go
100% Reserved capacity
100% Reserved capacity On-demand
100% Reserved capacity On-demand
Spot instances
Name-your-price
None
2. Ease of use is a pre-requisite 5 PRINCIPLES REPRODUCIBILITY
OF
http://headrush.typepad.com/creating_passionate_users/2005/10/getting_users_p.html
Help overcome the suck threshold
Easy to embrace and extend
Choose the right abstraction for the user
$ ec2-run-instances
$ starcluster start
None
None
Package and automate
Package and automate Amazon machine images, VM import
Package and automate Amazon machine images, VM import Deployment scripts,
CloudFormation, Chef, Puppet
Expert-as-a-service
None
None
1000 Genomes Cloud BioLinux
None
Your HiSeq data Illumina BaseSpace
DNA and RNA sequences Genomespace, Broad Institute at MIT
Data as a programmable resource
3. Reuse is as important as reproduction 5 PRINCIPLES REPRODUCIBILITY
OF
Seven Deadly sins of Bioinformatics: http://www.slideshare.net/dullhunk/the-seven-deadly-sins-of-bioinformatics
Seven Deadly sins of Bioinformatics: http://www.slideshare.net/dullhunk/the-seven-deadly-sins-of-bioinformatics
Infonauts are hackers
They have their own way of working
The ‘Big Red Button’
Fire and forget reproduction is a good first step, but
limits longer term value.
Monolithic, one-stop-shop
Work well for intended purpose
Challenging to install, dependency heavy
Inflexible
Embrace infonauts as hackers
Small things. Loosely coupled.
Easier to reuse
Easier to integrate
Scale out
Cancer drug discovery: 50,000 cores < $1000 an hour Schrödinger
and CycleServer
4. Build for collaboration 5 PRINCIPLES REPRODUCIBILITY OF
Workflows are memes
Reproduction is just the first step
Bill of materials: code, data, configuration, infrastructure
Full definition for reproduction
Utility computing provides a playground for data science
Code + AMI + custom datasets + public datasets +
databases + compute + result data
Code + AMI + custom datasets + public datasets +
databases + compute + result data
Code + AMI + custom datasets + public datasets +
databases + compute + result data
Code + AMI + custom datasets + public datasets +
databases + compute + result data
Package, automate, contribute.
Utility platform provides scale for production runs
5. Provenance is a first class object 5 PRINCIPLES REPRODUCIBILITY
OF
Versioning becomes really important
Especially in an active community
Doubly so with loosely coupled tools
Provenance metadata is a first class entity
Distributed provenance
5PRINCIPLES REPRODUCIBILITY OF
Remove constraints 5PRINCIPLES REPRODUCIBILITY OF
Accelerate science 5PRINCIPLES REPRODUCIBILITY OF
Chromosome 11 : ACTN3 : rs1815739
Chromosome X : rs6625163
Chromosome 19 : FUT2 : rs601338
+0.25 Chromosome 15 : rs2472297
Chromosome 2 : rs10427255
TYPE II Chromosome 10 : rs7903146
Chromosome 1 : rs4481887
Thank you aws.amazon.com @mza
[email protected]