Syntekabio

Discovery & Development

AI-driven DeepMatcher Drug Candidate Service

The Diagram shows Docker's extension to Syntekabio's AI-powered full drug discovery service solution that support almost all areas of drug discovery & development process from Hit discovery / Lead generation, and ADMET/PK to the pharmacogenomics biomarkers. It is based on Syntekabio’s powerful AI technologies (DeepMatcher®) that provide the best docking Hits/Leads through analysis of deep learning (3D-CNN) and MD simulation on targets. Docker is a virtualization software.

SCROLL DOWN

DDCS : HIT to Preclinical Candidate in 2 YEARS

Hit

Hit Screening

DeepMatcher - Hit
Lead

Lead Generation

DeepMatcher - Lead
Candidate

ADMET/ PK

DeepMatcher - ADMET/PK
Preclinical

GLP-Tox

FDA Registered Facility (CRO)
Precision Medicine

Clinical

PGx Biomarkers for Drug Labeling

Immunogenic Neoantigen Prediction

Cancer/CNS Patient Stratification SNPs

CYP typing, HLA typing & KIP typing

A-Z Package

Hit-Discovery, Assay Validation, Lead Generation & Preclinical Study
Client-Centric

Minimal Risk *Maximum Customization
In-House & Global
CRO Cooperation

Discovery & Development

Infra-structure

Fully Customized Packages & Services to Ensure High Success Rate

Bio-Supercomputer Capacity for Bio-Research, Medicine & Pharmacology
Experiences

15+ Years Experience on Machine-Learning + Structural biology + Supercomputing

AI & GPT Accelerates Screening Flows and Calculations and increase success rate

30+ Affiliations & Partnerships with Certified Global CRO Expert Groups
Owner-finance

7.5% Down Payment of Total $cost to Start & Installments for Pay As You Go Model

Flexible Timeline & Milestone to Match Client’s Finance & Convenience
All in one

From Start-to-Finish A-Z Services = Identify Preclinical Drug Candidate in 2 YEARS & Support IND Preparation

HIT to Preclinical Candidate in 2 YEARS

Bio-supercomputer AI-data center with 10,000 servers (in progress)
CLOUD based ABC serves best-in-class solutions for AI drug discovery

>5~8 yrs+>10yrs, 1 approved drug

Drug Discovery

Preclinicial

Clinic Development

1 yr

2 yr

5 yr

Hit Id

Hit to Lead

Lead Opt.

Preclinicial

1 yr

Hit

Lead-gen

ADMET/Pk

STB CLOUD + DeepMatcher

Hit

All in 50 Days Form Inial Feasbility
Test to Done Screening

In-vitro assat 100 days, Start to Lead :
Entry Point within 150 Days

Lead-gen

Lead gen.
in 210 days

ADMET/PK

1 Year Period:
form leads to
ADMET Screening

DeepMatcher (Preclinical) Drug Candidate : DDC framework

Auto-Hit Screening

1B purchasable Cpnd, screening
1M Physical docking
1K 3D-CNN binding tunning
1K Auto-MD Simulatio
In-vitro validation
Al-Hits <1μM

Finding Grooves (sub-pocket)
20K R-group Substitution
Spatial Occupancy Filtering
1K Auto-MD Simulation
Iln-vitro/In-vivo validation
Al-Leads <10nM

Auto-Lead Generation

Auto-ADMET/
PK Prediction

BERT pre- trained Model
Multi-tasking ADMET profiling
3D-CNN Toxicity prediction
PK prediction
In-vitro/In-vivo validation
Preclinical candidate

Seeking Ultimate Success

AI-POWERED auto-HIT/Lead discovery: DDCS, as successful loop platform

1Auto-HIT Discovery
- Purchasable 1B Chemical Library
- Physical-3D Docking (1M)
- 3D-CNN Pose Tunning (1K)
- Auto-MD Simulation Fine Tunning (1K)
2Auto-Lead Gen.
- Sub-pocket & Week bond Screening
- Substitution of 20K R-group to Scaffold
- Free Energy Perturbation in the Sub-pocket
- Auto-MD Simulation Fine Tunning (1K)
3Auto-ADMET.PK Scr.
- Solubility,Intestinal Absorption, Metabolism, PPB, BBB
- Liver injury, Off target toxicity (NR,HLA adduct, Signaling), PK, etc.

Deep Learning : Strength in docking & best pose generation similar to real 3D structure
Auto-MD simulation : Strength in binding energy calculations close to experimental values
GPT : Streng thin virtual screening by cognitive learning based on big data and extreme hardware

Chemical Library
screening
3D-CNN Binding Tuning
R-group substitution
Auto-MD Simulation
BERT – ADMET/PK screening
3bmGPT - key residue analysis

Service Process

Step
Items
Core Technology
TPP
Duration

Hit Discovery

AI-Hit discovery

AI-Hit purchase

Hit confirm

DMC-Hit®

IC50 < 1μM~8 Hits

~2m
~3m
Lead Generation

AI-LEADgeneration

Compound synthesis & Lead confirm

Go-No Go Decision (1st Year)

DMC-Lead®

IC50 < 10nM ~4 Leads

~7m
ADMET/PK(Preclinical Candidate)

AI-ADMETgeneration

Compound synthesis

ADMETvalidation

PKvalidation in mice (PO)

Success-Failure Decision (2nd year)

DMC-ADMET/
PK predictor®

Decent ADMET
/PK Profiles
~1 Candidate

~12m

Cost Breakdown

Unit
Hit Discovery
Lead Generation
ADMET/PK
Total

Hardware/computation
(single service only*)

100 Servers / 3Month

100 Servers / 3Month

100 Servers / 3Month

-

$150K

$150K

$150K

-
Chemical
Purchase/Synthesis/Assay
(Validationsize)

~200 chemicals

~200 chemicals

~200 chemicals

-

$20K ($100/chemical/
purchase)

$300K (10mg/chemical/
synthesis)

$200K (~150mg/
chemical/
synthesis)

-

$20K ($100/chemical/
purchase)

$7.5K ($150/chemical)

$400K (~150mg/
chemical/
synthesis)

-
Estimated Cost sum

$200K

$457K

$750K

Total

$1.4 Million
BigPharma Traditional R&D Cost④

$1.0M

$12.5M

$12.5M

Total

$13.5M

Medium.com/EvinceBio, 2017 report

References:

Syntekabio’s Hit to PreclinicalCandidate on Cloud Computing, BioPharmaTrend. Vol4.2023
How Syntekabio’s Supercomputing Enables AI-driven Drug Discovery, BioPharmaTrend, Vol3. 2023
The Power of STB CLOUD in Remote AI Drug Discovery, BioPharmaTrend, Vol2. 2023
The State of A.I. Drug Discovery and Its Future:Small Molecules, Vaccines, and Antibodies, BioPharmaTrend, Vol1. 2023

DDC AI-driven DeepMatcher
(preclinical) Drug
Candidate Service

DDCS : HIT to Preclinical Candidate in 2 YEARS

The Diagram shows Docker's extension to Syntekabio's AI-powered full drug discovery service solution that support almost all areas of drug discovery & development process from Hit discovery / Lead generation, and ADMET/PK to the pharmacogenomics biomarkers. It is based on Syntekabio’s powerful AI technologies (DeepMatcher®) that provide the best docking Hits/Leads through analysis of deep learning (3D-CNN)and MD simulation on targets. Docker is a virtualization software.

Hit

Hit Screening

DeepMatcher - Hit
Lead

Lead Generation

DeepMatcher - Lead
Candidate

ADMET/ PK

DeepMatcher - ADMET/PK
Preclinical

GLP-Tox

FDA Registered Facility (CRO)
Precision Medicine

Clinical

PGx Biomarkers for Drug Labeling

Immunogenic Neoantigen Prediction

Cancer/CNS Patient Stratification SNPs

CYP typing, HLA typing & KIP typing

A-Z Package

Hit-Discovery, Assay Validation, Lead Generation & Preclinical Study
Client-Centric

Minimal Risk * Maximum Customization
In-House & Global
CRO Cooperation

Discovery & Development

Infra-structure

Fully Customized Packages & Services to Ensure High Success Rate

Bio-Supercomputer Capacity forBio-Research, Medicine & Pharmacology
Experiences

15+ Years Experience onMachine-Learning + Structural biology + Supercomputing

AI & GPT Accelerates Screening Flows and Calculations and increase success rate

30+ Affiliations & Partnerships with Certified Global CRO Expert Groups
Owner-finance

7.5% Down Payment of Total $cost toStart & Installments for Pay As You Go Model

Flexible Timeline & Milestone to Match Client’s Finance & Convenience
All in one

From Start-to-Finish A-Z Services = Identify Preclinical Drug Candidate in 2YEARS & Support IND Preparation

HIT to Preclinical Candidate in 2 YEARS

Bio-supercomputer AI-data centerwith 10,000 servers (in progress)
CLOUD based ABC servesbest-in-class solutions for AI drug discovery

>5~8 yrs+>10yrs, 1 approved drug

Drug Discovery

Preclinicial

Clinic Development

1 yr

2 yr

5 yr

Hit Id

Hit to Lead

Lead Opt.

Preclinicial

1 yr

Hit

Lead-gen

ADMET/Pk

STB CLOUD + DeepMatcher

Hit

All in 50 Days Form Inial Feasbility
Test to Done Screening

In-vitro assat 100 days, Start to Lead :
Entry Point within 150 Days

Lead-gen

Lead gen.
in 210 days

ADMET/PK

1 Year Period:
form leads to
ADMET Screening

DeepMatcher (Preclinical) Drug Candidate : DDC framework

Auto-Hit Screening

1B purchasable Cpnd, screening
1M Physical docking
1K 3D-CNN binding tunning
1K Auto-MD Simulatio
In-vitro validation
Al-Hits <1μM

Finding Grooves (sub-pocket)
20K R-group Substitution
Spatial Occupancy Filtering
1K Auto-MD Simulation
Iln-vitro/In-vivo validation
Al-Leads <10nM

Auto-Lead Generation

Auto-ADMET/
PK Prediction

BERT pre- trained Model
Multi-tasking ADMET profiling
3D-CNN Toxicity prediction
PK prediction
In-vitro/In-vivo validation
Preclinical candidate

Seeking Ultimate Success

AI-POWERED auto-HIT/Lead discovery: DDCS, as successful loop platform

1Auto-HIT Discovery
- Purchasable 1B Chemical Library
- Physical-3D Docking (1M)
- 3D-CNN Pose Tunning (1K)
- Auto-MD Simulation Fine Tunning (1K)
2Auto-Lead Gen.
- Sub-pocket & Week bond Screening
- Substitution of 20K R-group to Scaffold
- Free Energy Perturbation in the Sub-pocket
- Auto-MD Simulation Fine Tunning (1K)
3Auto-ADMET.PK Scr.
- Solubility,Intestinal Absorption, Metabolism, PPB, BBB
- Liver injury, Off target toxicity (NR,HLA adduct, Signaling), PK, etc.

Deep Learning : Strength in docking & best pose generation similar to real 3Dstructure
Auto-MD simulation : Strength in binding energy calculations close to experimentalvalues
GPT : Strengthin virtual screening by cognitive learning based on big data and extremehardware

Chemical Library
screening
3D-CNN Binding Tuning
R-group substitution
Auto-MD Simulation
BERT – ADMET/PK screening
3bmGPT - key residue analysis

Service Process

Step
Items
Core Technology
TPP
Duration

Hit Discovery

AI-Hit discovery

AI-Hit purchase

Hit confirm

DMC-Hit®

IC50 < 1μM~8 Hits

~2m
~3m
Lead Generation

AI-LEADgeneration

Compound synthesis & Lead confirm

Go-No Go Decision (1st Year)

DMC-Lead®

IC50 < 10nM ~4 Leads

~7m
ADMET/PK(Preclinical Candidate)

AI-ADMETgeneration

Compound synthesis

ADMETvalidation

PKvalidation in mice (PO)

Success-Failure Decision (2nd year)

DMC-ADMET/
PK predictor®

Decent ADMET
/PK Profiles
~1 Candidate

~12m

Cost Breakdown

Unit
Hit Discovery
Lead Generation
ADMET/PK
Total

Hardware/computation
(single service only*)

100 Servers / 3Month

100 Servers / 3Month

100 Servers / 3Month

-

$150K

$150K

$150K

-
Chemical
Purchase/Synthesis/Assay
(Validationsize)

~200 chemicals

~200 chemicals

~200 chemicals

-

$20K ($100/chemical/
purchase)

$300K (10mg/chemical/
synthesis)

$200K (~150mg/
chemical/
synthesis)

-

$20K ($100/chemical/
purchase)

$7.5K ($150/chemical)

$400K (~150mg/
chemical/
synthesis)

-
Estimated Cost sum

$200K

$457K

$750K

Total

$1.4 Million
BigPharma Traditional R&D Cost④

$1.0M

$12.5M

$12.5M

Total

$13.5M

Medium.com/EvinceBio, 2017 report

References:

Syntekabio’s Hit to PreclinicalCandidate on Cloud Computing, BioPharmaTrend. Vol4.2023
How Syntekabio’s Supercomputing Enables AI-driven Drug Discovery, BioPharmaTrend, Vol3. 2023
The Power of STB CLOUD in Remote AI Drug Discovery, BioPharmaTrend, Vol2. 2023
The State of A.I. Drug Discovery and Its Future:Small Molecules, Vaccines, and Antibodies, BioPharmaTrend, Vol1. 2023

Quickly experience the service

AI-driven DeepMatcher Drug Candidate Service

DeepMatcher (Preclinical) Drug Candidate : DDC framework

Service Process

Cost Breakdown

DeepMatcher (Preclinical) Drug Candidate : DDC framework

Service Process

Cost Breakdown

send us your inquiry