Next generation genomic analysis for next generation healthcare - - PowerPoint PPT Presentation

next generation genomic analysis for next generation
SMART_READER_LITE
LIVE PREVIEW

Next generation genomic analysis for next generation healthcare - - PowerPoint PPT Presentation

Next generation genomic analysis for next generation healthcare GENOMIC SEQUENCING | RAPIDLY FALLING PRICES 2003 2018 In 1990, Congress established funding for the Human Genome Project. The project ended up costing about $2.7 billion in FY


slide-1
SLIDE 1

Next generation genomic analysis for next generation healthcare

slide-2
SLIDE 2

GENOMIC SEQUENCING | RAPIDLY FALLING PRICES

2

In 1990, Congress established funding for the Human Genome

  • Project. The project ended up

costing about $2.7 billion in FY 1991 dollars

2003 2018

slide-3
SLIDE 3

GENOMIC SEQUENCING | RAPIDLY FALLING PRICES

3

Now

slide-4
SLIDE 4

SEQUENCING IS CHEAP| INTERPRETING THE DATA IS NOT

4

GENERATE RAW DATA
 (Sequencing)

SOURCE: The Economist, 2014, SeqOne 2016 (Not to scale)

Cost & Time / Patient Time 2015 $2B

ANALYSE DATA

Interpretation now more expensive than sequencing

€300

2015

slide-5
SLIDE 5

SeqOne - Business overview / CONFIDENTIAL

ANALYSIS COMPLEXITY | BEYOND GENE-PANEL

5

* NGS: New Generation Sequencing

10%

Legacy gene-panel analysis

Gene-panel CNV & SNV Exome SNV & CNV RNA differential expression Exome SV

Analysis complexity Number of patients

90%

Requires richer genomic analysis enabled by NGS* data

A single optimised application for each specific analysis need (eg. Vidgil for VDJ re-arrangements in RNA)

SEQONE

AVAILABLE TOOLS

Etc….

slide-6
SLIDE 6

SeqOne - Business overview / CONFIDENTIAL

6

Scientists are expecting as many as 1 billion people to have their genomes sequenced by 2025.

Washington Post 2017

PATIENT DEMAND | REQUIRES MASSIVE PROCESS SCALING

slide-7
SLIDE 7

SeqOne - Business overview / CONFIDENTIAL

CURRENT PROCESS | FRAGMENTED, SLOW, EXPENSIVE

7

Transfer Compute / Store

Sequencing Bioinformatics (identify anomalies) Doctor

ANALYSIS AND INTERPRETATION

Expertise-intensive and expensive

SEQUENCING

Automated & affordable

Iteration Iteration

Specific genes
 (today)

Raw data Start of process

4 weeks - 6 months

Genomics expert (interpretation) Infrastructure (data mgt.)

slide-8
SLIDE 8

8

SeqOne vision

A solution that makes genomic testing at high volume as easy, accessible and accurate as any medical test your doctor can order today

slide-9
SLIDE 9

SeqOne -CONFIDENTIAL

SEQONE TEAM | GENOMIC ANALYSIS EXPERTS

  • N. Phillipe - CEO / CSO: 10 years experience in NGS

genomic analysis, Author open-source RNA-Seq analysis tools, > 20 bioinformatic academic papers. Many international collaborations G Buwalda - CTO: 15 years secure high volume transactional computing (fintech) 
 JM Holder - CCO: 25 years experience in international business, ex Apple, Palm, Packet Video. Founder of two funded startups. British citizen. Current team of 10 specialists in bioinformatics, biology and CIS. Winner of numerous prizes (iLab, ARC, etc.)

9

Paris Montpellier

FRANCE

CTO CCO CEO

slide-10
SLIDE 10

SeqOne -CONFIDENTIAL

SEQONE | GENOMIC ANALYSIS EXPERTS

10

Paris Montpellier

FRANCE

CEO CTO

CCO

Project launch 2014 / Incorporated 2017 Current team of 10 Winner ARC H. Stark, iLab Emergence / Creadev €950 funding through SATT and subsidies

  • Dr. N Philippe


10 years, Bioinformatics, RNA- Seq specialist

  • G. Buwalda


15 years, secure, big- data, high availability,

  • pen systems

J-M Holder
 25 years international biz / startups, ex Apple, British

CSO

  • Dr. J Audoux


5 Years, Computer scientist & Bioinformatics from RNA to DNA

slide-11
SLIDE 11

11

slide-12
SLIDE 12

SeqOne | GENOMIC ANALYSIS SIMPLIFIED

12

Supports a wide range of genomic tests AI-driven decision support improves speed and accuracy of results Scales to meet tomorrow’s patient volumes

A SaaS solution that makes genomic testing as easy, accessible and accurate as a blood test

“SeqOne allows us to perform miracles; we can handle four times the number of patients with the same resources”

  • Pr. Lebre - CHU Reims
slide-13
SLIDE 13

SeqOne - Business overview / CONFIDENTIAL

13

USABILITY | EASY-TO-USE AI-DRIVEN PLATFORM

Coefficient of variation

Intuitive UX AI ranked variant short-list Artefact removal App-specific visualization

Truly Short “Shortlists”

slide-14
SLIDE 14

SeqOne - Business overview / CONFIDENTIAL

CHU / PRIVATE OTHER RARE DISEASES BASIC CANCER APPS

SUPPORT FOR ALL TYPES OF ANALYSIS | APP CATALOG

14

CORE APPS

Somatic Mono

(single sample)

Somatic Duo

(healthy / tumour)

Trio

(family)

Mono

(single sample)

PREMIUM APPS

ADVANCED CANCER APPS

ONCOMPASS

(Cancer tracking)

INCA -Sardine

(Bias removal)

CNV

(DNA copies)

Other

QC

(Sample quality)

Mono + Phenotips

(CHUGA / Gene42)

Private

(blocked)

Fusion genes

(RNA analysis)

+

GeneExp RNA (Multipurpose)

slide-15
SLIDE 15

SeqOne - Business overview / CONFIDENTIAL

SCALABILITY | CLOUD/LOCAL INFRASTRUCTURE

15

SeqOne Cloud Local SeqOne Instance External Data Local
 Interpretation Environment & 
 Data store SeqOne Cloud External Data Cloud User

slide-16
SLIDE 16

SeqOne - Business overview / CONFIDENTIAL

REGULATORY | QUALITY, TRACEABILITY, SECURITY

Platform certification (In-vitro diagnosis CE Mark) All data encrypted and hosted on medical grade secure servers Each data-file, variant and action on the platform logged Versioning of platform, apps and databases to ensure reproducibility Entire process managed on SeqOne from sequencer to final report facilitating lab certification (ISO 15189, COFRAC, etc.)

16

slide-17
SLIDE 17

SeqOne - Business overview / CONFIDENTIAL

ARCHITECTURE | MODULAR, OPTIMISED, DISTRIBUTED

17

CLOUD INSTANCE VIRTUALISED, SECURE MODULAR STORAGE / COMPUTE EASY-TO-USE INTERPRETATION ENVIRONMENT PUBLIC ANNOTATIONS CACHE OPTIMISED SOMATIC BIOINFORMATIC FRAMEWORK LOCAL 1 LOCAL 2 PRIVATE STORE 1 PRIVATE STORE 2 OPTIMISED MENDELIAN BIOINFORMATIC FRAMEWORK MACHINE LEARNING / AI ENGINE GENOMIC EVENTS DB

APPS

slide-18
SLIDE 18

SeqOne - Business overview / CONFIDENTIAL

GENOMIC ANALYSIS | MEDICAL APPLICATIONS

18

Rare diseases Pre-natal, non- invasive testing Cancer Complex disease

$4B 


15% CAGR

(2017 est.)*

* SOURCE: Bioinformatics Market by Product & Service - Global Forecast to 2023", published by MarketsandMarkets™

slide-19
SLIDE 19

SeqOne - Business overview / CONFIDENTIAL

RARE DISEASES | KEY FACTS

Not that rare: 6-8% of Euro pop. impacted (estimated 350M WW) Serious: 30% of patients die before the age of 5 (source Eurodis) Genomic: 80% linked to genetic causes 6 out of 7 patients not diagnosed (15% diagnostic yield)

19

slide-20
SLIDE 20

SeqOne - Business overview / CONFIDENTIAL

IMPROVING DIAGNOSTIC YIELD | NGS EXOME

NGS delivers “high-resolution” exome genomic data Exomes increase diagnostic yield from 15% to 35%-50% BUT Exome analysis costs higher

  • >3X sequencing (must sequence

entire family)

  • Longer time required to analyse

20

«Exome sequencing means that genomic diseases can be diagnosed more quickly »

slide-21
SLIDE 21

SeqOne - Business overview / CONFIDENTIAL

EXOMES TOO COSTLY | EXOME FAMILY TRIO VS GENE PANEL

21

1250 2500 3750 5000 Gene panel (SeqOne est.) Exome Trio

€1 543 €600 2 194 € 200 € 408 € 550 €

Preparation Sequencing Analysis

€ 1.350 € 4.145

slide-22
SLIDE 22

HELPING RD DIAGNOSTIC WITH AI SOLUTIONS

Current Clinical interpretation strategies involves

  • Hard-filtering
  • trio-based filtering to shorten variant list (de novo, het-composite)
  • long manual review into a Excel Sheet

SeqOne introduce a Phenotype-based AI ranking strategy to :

  • Reduce interpretation time made by the expert
  • Cost-reduction: Index-case sequencing only
  • Augment diagnostic veld, through IA vs Hard-filtering solution

22

slide-23
SLIDE 23

SeqOne - Business overview / CONFIDENTIAL

DATA CONSISTENCY | RARE DISEASE TRIAL

30 Exomes for learning - 300 Exomes for testing Data from 8 CHU’s, heterogenous phenotyping & bioinformatics Rankings - Data consistency impacts results:

23

First 75%

4-20 15 % 2-3 30 % First 56 %

learning dataset: all in top 5 Testing datastet: all in top 20

Median 1 Mean 5

slide-24
SLIDE 24

SeqOne - Business overview / CONFIDENTIAL

OUTCOMES | RARE DISEASE TRIAL

24

Causal variants in top 5 on average < 30mins per diag 5% new diagnoses vs control 45%-50% diag yield / single exome
 (2.5 X panel perf.)

+

slide-25
SLIDE 25

SeqOne - Business overview / CONFIDENTIAL

HOME MARKET | TRACTION IN RARE DISEASES

25

Brest Lille Creteil Reims Strasbourg Grenoble Montpellier Lyon Clermond-Ferrand Rennes CHU User Collaboration Private sector Paris Nantes Rouen Dijon

slide-26
SLIDE 26

26

SEQONE | CONCLUSION

The simplest, most efficient way to interpret NGS data at scale in clinical settings

Dynamic exploration Fast, accurate results Optimized bioinformatics Secure upload and storage

slide-27
SLIDE 27

27