2. Bio startups that operate like software startups
The emerging macro trend: Bio 2.0
Bio in 2015 is like software in 2005
Software is eating Bio
Cloud biology
emergence of low
CapEx startups
Software at the center
Machine learning,
cloud computing
Minimize FDA Risk
DTC, digital health,
consumer genomics, etc
3. Storage cost-performance and computing cost-performance
Moore’s law: cost of storage, compute ⇒ zero
0.01
0.1
1.
10.
100.
1000.
1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
Dollars($)
Compute cost ($ per 1 million transistors)
Storage cost ($ per gigabyte)
4. Silicon content: $20-22
NAND Flash
$18-20
Display
$15-17
Applications Processor
$8-10
DRAM
$10-13
Baseband
$4-5
RF and FEM
$3-4
Power
Amplifier
$3-4
PMIC
$3-4
Combo-chip
(WIFI/BT/FM)
$3-4
Touch
Controller
$3-4
GPS
$3-4
Image
Sensors
Silicon content: $9-10
Camera Module
$7-11
Touch Panel
$5-6
Battery (Li Polymer)
$3-5
HDI PCB
$1-2
Camera Lens
$2-3
Gyroscope/Accelerometer
$0.70-0.80
Audio
Codecs
$0.50-0.60
Speaker IC
Substrate
$1-2
LCD Drive IC
$0.70-0.80
MEMs Microphone
$1.00-1.50
LCD Drive IC
Moore’s law: cost of storage, compute ⇒ zero
Source: Nomura Securities, Gartner 2013 report.
5. Side benefits of Moore’s law: cost of sensors ⇒ zero
Source: Qualcomm
IntegratedSensors,UserExperiences
Ambient Light
Accelerometer
Magnetometer
Ambient Light
Accelerometer
Magnetometer
Ambient Light
Accelerometer
Magnetometer
Ambient Light
Accelerometer
Magnetometer
Ambient Light
Accelerometer
Magnetometer
Gyroscope
Proximity
Gyroscope
Proximity
Gyroscope
Proximity
Pressure
RGB
Pressure
RGB
Pressure
RGB
Gyroscope
Proximity
Temperature
Humidity
Hall Effect
Temperature
Humidity
Hall Effect
Heart Rate
Fingerprint
2010 2011 2012 2013 2014 2015+
GALAXY 1
GALAXY S2
GALAXY S3
GALAXY S4
GALAXY S5
6. Beyond Moore’s law: cost of sequencing ⇒ zero
Source: Nature, 2014
Cost of genome
sequencing.
Next generation
sequencers enter
the market.
Moore’s law for
computing costs.
The price of
sequencing a whole
human genome hovers
around $5,000 and is
expected to drop even
lower.
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
10,000
1,000
100
10
1
Cost(thousandsUS$)
7. Putting these trends together
This is disrupting traditional biotech
= SOFTWARE IS EATING BIO
Compute Sensors Biology
+ +
8. Bio 2.0 is not Biotech
BIOTECH STARTUPS SOFTWARE STARTUPS
Subject Uncontrollable organisms Perfectly determinate code
Environment Poorly understood, natural Well understood, artificial
Approach Indefinite, random Definite, engineering
Regulation Heavily regulated Basically unregulated
Cost Expensive ( > $1B per drug) Cheap (a little seed money)
Team High-salaried, unaligned lab drones Committed entrepreneurial hackers
& BIO 2.0
From Peter Thiel’s Zero to One
12. Slow and expensive to develop
Toxic side effects
FDA regulated
Traditional therapeutics
IMAGE: Champlax
13. Title Text
Subtitle text
Digital therapeutics for chronic
disease
Mobile enables traction, patient
success, and a new model for
insurance companies
Digital health & digital therapeutics
IMAGE: Omada Health
14. Large capital outlay
Poor reproducibility
Empirically driven
doesn’t scale
The four horsemen of Eroom’s law
Traditional biology
15. Title Text
Subtitle text
Biology that works like programming:
Cloud experiments run through software
Cloud biology
IMAGE: Emerald Therapeutics
18. New a16z fund for Bio 2.0
Vijay Pande new GP:
uniquely suited for Bio 2.0
a16z approach
applied to Bio
Bio 2.0 software +
cloud bio, w/o
traditional FDA risk
20. Vijay Pande:
Bio2.0 background
Chemistry
Camille and Henry Dreyfus Professor, Stanford Univ.
Thomas Kuhn Paradigm Shift Award, American Chemical
Society Teacher-Scholar Award, Dreyfus Foundation
Physics
AB Princeton University, 1992, PhD MIT, 1995 | Michael and
Kate Bárány Award for Young Investigators, Biophysical
Society Fellow, American Physical Society
Computer Science:
Founder, Folding@home Distributed Computing | MIT TR100
Guinness World Record, Folding@home First to a Petaflop
Netxplorateur of the Year
Biology
Chair, Biophysics, Stanford University Medical School
Delano Award for Computational Biosciences, American Soc.
for Biochem. and Molecular Biology | Irving Sigal Young
Investigator Award, Protein Society
21. Title Text
Subtitle text
Crowd sourced computational
biology for biomedicine
Anticipated the future
Cloud before Loudcloud
GPU computing before CUDA
Crowdsourcing before wikipedia
Founding Director, Folding@home
Approximately 2,000,000
people have donated
computer time.
More processing power than
Amazon Web Services,
Guinness World Record
FOLDING@HOME
22. Title Text
Subtitle text
$900M
Next generation approaches to
therapeutics for infectious disease
Computation at its heart
Regulatory Innovations: drug repurposing
Co-founder, Globavir BioSciences
23. Vijay Pande – Bio & IT entrepreneur
ADVISOR TO STARTUPS IN BIO
Acumen Pharmaceuticals (Alzheimer’s
Disease)
Counsyl (Carrier testing)
Globavir (Infectious Disease)
EMPLOYEE #1 AT NAUGHTY DOG
Started at 15 years old
writing computer games
Naughty Dog later
sold to Sony
ADVISOR TO STARTUPS IN IT
Clearspeed (high speed compute hardware)
Discovery Engine (Semantic search)
Numerate (Computational Drug Design)
Omnipod (Software for pharma)
OpenEye (Software for pharma)
Pharmix (Computational Drug Design)
Protein Mechanics (molecular simulation)
Schrodinger (Software for pharma)
Stack IQ (Cloud software stack)
Editor's Notes
Powerful new methods put software at the center
Machine Learning, Large-scale data, Cloud computing
Emergence of a “Cloud Bio” infrastructure
Analogs to cloud computing: low CapEx startups in Bio
Approaches that seek to avoid FDA risk
DTC, consumer genomics, etc
Powerful new methods put software at the center
Machine Learning, Large-scale data, Cloud computing
Emergence of a “Cloud Bio” infrastructure
Analogs to cloud computing: low CapEx startups in Bio
Approaches that seek to avoid FDA risk
DTC, consumer genomics, etc
Cloud biology
Biology is following in IT’s footsteps with its own cloud experimental infrastructure
Ecosystem gives ability to startup with minimal investment
Computational medicine
Genomic advances enabling personalized medicine
Machine learning enabling smart drug discovery
Digital Health
Mobile phone and software enabling broad population management tools
Cloud biology
Biology is following in IT’s footsteps with its own cloud experimental infrastructure
Ecosystem gives ability to startup with minimal investment
Computational medicine
Genomic advances enabling personalized medicine
Machine learning enabling smart drug discovery
Digital Health
Mobile phone and software enabling broad population management tools
Software can positively impact our health, wellness, and lives
Mobile leads to more natural human connections, sensors, reminders
Quantified self + Cloud brings everything together
Smartphones enable better population management tools
Smartphones and notifications allow engagement with healthcare apps
Smartphone hardware can be used for diagnostic purposes or to power medical add-ons
Teledermatology use phone camera to send pictures of skin conditions
Decentralization as a general theme
Bringing the patient to the center of the decision making process
Cloud biology
Biology is following in IT’s footsteps with its own cloud experimental infrastructure
Ecosystem gives ability to startup with minimal investment
Computational medicine
Genomic advances enabling personalized medicine
Machine learning enabling smart drug discovery
Digital Health
Mobile phone and software enabling broad population management tools
Hardware: Cloud Experiments
equivalents to data centers for experiments: low startup CapEx
Cloud labs can use commodity hardware from mobile supply chain
Software
Machine Learning/Statistics powerfully complement biology
Open source tools create an established tool chain
People
New generation of students who know biology and can code
Coding is more than programming — it’s the way to break down challenging problems, applicable to biology broadly
Cloud biology
Biology is following in IT’s footsteps with its own cloud experimental infrastructure
Ecosystem gives ability to startup with minimal investment
Computational medicine
Genomic advances enabling personalized medicine
Machine learning enabling smart drug discovery
Digital Health
Mobile phone and software enabling broad population management tools
Consumer: consumer genomics: eg, actionable choices
Clinical
personalized medicine: eg, use genomics to predict the best drugs for cancer patients
advanced computational imaging: eg coupling machine learning with advances in imaging hardware
clinical genomics & testing: eg disrupting existing clinical tests with genomic approaches that are cheaper and open new doors
B2B: new tools to accelerate others, building out the infrastructure
Powerful new methods put software at the center
Machine Learning, Large-scale data, Cloud computing
Emergence of a “Cloud Bio” infrastructure
Analogs to cloud computing: low CapEx startups in Bio
Approaches that seek to avoid FDA risk
DTC, consumer genomics, etc
Software can positively impact our health, wellness, and lives
Mobile leads to more natural human connections, sensors, reminders
Quantified self + Cloud brings everything together
Smartphones enable better population management tools
Smartphones and notifications allow engagement with healthcare apps
Smartphone hardware can be used for diagnostic purposes or to power medical add-ons
Teledermatology use phone camera to send pictures of skin conditions
Decentralization as a general theme
Bringing the patient to the center of the decision making process