[Volume 15. Drug Discovery AI Innovation - Schrödinger Inc.]

schrodinger Quick shape workflow

Schrödinger: Leading Next-Generation Platform Through the Convergence of Physics-Based Computational Chemistry and AI

Schrödinger is a global leader in accelerating drug discovery by combining physics and machine learning. Since its founding in 1990, the company has accumulated over 30 years of computational chemistry expertise and leverages cutting-edge AI technology to dramatically reduce the time and cost of drug development.

As a data analytics and software development professional, the biggest challenge in traditional drug discovery has been accurately predicting complex molecular interactions. Conventional AI/ML tools often increase technical complexity without providing fundamental solutions. However, Schrödinger's approach is completely different.

Research Overview

Company: Schrödinger, Inc. (Nasdaq: SDGR)

Website: https://www.schrodinger.com

Domain: Physics-Based Molecular Simulation, AI Drug Discovery Platform

Core Solutions: Physics-Based Computational Platform, LiveDesign, Maestro, FEP+

Core Technology: Molecular Dynamics Simulation, Free Energy Perturbation (FEP+), Machine Learning, Cloud Computing

Headquarters: New York, NY, USA

Korean Office: Pangyo Innovalley A, Gyeonggi-do, Korea

Founded: 1990 (Named after Nobel Physics Prize winner Erwin Schrödinger)

CEO: Ramy Farid (PhD Chemistry, Caltech, former Rutgers University Professor)

IPO: February 2020 Nasdaq listing (Ticker: SDGR)

Employees: Approximately 800 (Connected with customers and collaborators in over 70 countries)

2024 Revenue: Total revenue approximately $238 million (Software revenue increased 21%)

Key Investors: Baron Capital, Baillie Gifford, Wellington Management

Strategic Partnerships:

• All Global Top 20 Pharmaceutical Companies

• Bristol Myers Squibb (Up to $2.7 billion contract)

• Takeda, Novartis, Gilead Sciences, Lilly

• Google Cloud, AWS, Microsoft Azure

Schrödinger's Innovative Approach: Fusion of Physics-Based AI and Machine Learning

1. Core Technology of Physics-Based Computational Platform

Schrödinger's differentiated strength lies in its hybrid approach combining accurate calculations based on physical laws with the speed and scalability of machine learning.

Core Technology Stack:

Molecular Dynamics Simulation:

• Simulates molecular interactions in real biological environments at the atomic level

• Large-scale parallel processing through GPU-based high-performance computing

• Unlimited scalability secured through strategic partnership with Google Cloud

Free Energy Perturbation (FEP+):

• Predicts binding affinity of drug candidates with chemical accuracy

• Over 95% accuracy in predicting drug-protein binding compared to experiments

• 100-1000 times faster than traditional experimental methods

Active Learning System:

• Uses physics-based calculation results as training data

• Screens billions of compounds at 0.1% of the cost

• Achieves approximately 70% of the same performance while dramatically reducing costs

2. AI Integration Strategy: Physics-Driven Machine Learning

CEO Ramy Farid describes it as "Physics is the cake, and machine learning is just the cherry on top." This well represents Schrödinger's unique AI philosophy.

Latest De Novo Design Workflow (Released May 2025):

AutoDesigner Innovation Cases:

• EGFR Discovery Project: Explored 23 billion new chemical structures, discovered 4 novel scaffolds in just 6 days

• WEE1 Inhibitor Development: Generated completely new chemotypes with >10,000x selectivity over PLK1

• AutoDesigner LinkerDesign: Automatic generation and evaluation of billions of potential linkers between molecular fragments

AI Utilization Methods:

Physics-First Approach: CEO Ramy Farid describes it as "Physics is the cake, and machine learning is just the cherry on top," emphasizing verifiable AI systems that ensure physics-based accuracy over black-box solutions.

Current LLM Integration Status:

• Currently limited direct integration with general-purpose LLMs (GPT, Claude, Gemini)

• Instead, develops domain-specific AI models internally

• Builds verifiable AI systems that ensure physics-based accuracy

3. Unlimited Scalability Based on Cloud

Google Cloud Strategic Partnership:

Computing Resources:

• Approximately 1 GPU-day of computation required per molecule (100-200 CPU-days on CPU basis)

• Stable network connections for simultaneous simulation of thousands of molecules

• Prevention of job interruption due to transient network instability

COVID-19 Response Case:

• Collaboration with Takeda, Novartis, Gilead Sciences to accelerate therapeutic development

• Goal to shorten traditional drug discovery from 5-6 years to 2-3 years

4. Practical Implementation and Achievements

Schrödinger's technology has delivered concrete results across multiple therapeutic areas and partnerships.

Major Success Stories:

Bristol Myers Squibb Collaboration:

• Signed in 2020, total $2.7 billion scale ($55 million upfront)

• Multi-target drug discovery in oncology, immunology, and neurology

• Advancing HIF-2α (kidney cancer), SOS1-KRAS (oncology) programs

Takeda Partnership:

• Development of TYK2 inhibitor TAK-279 through Nimbus Therapeutics

• Takeda's $1.1 billion acquisition resulted in $111 million revenue for Schrödinger

Gilead Sciences Collaboration:

• Licensed ACC inhibitor program for $1.2 billion

• Schrödinger received total $46 million revenue in 2016-2017

Competitive Analysis: Schrödinger vs Major Competitors

Aspect	Schrödinger	OpenEye Scientific	Cresset Group	BioSolveIT
Core Approach	Physics-Based + AI Hybrid	Computational Chemistry Toolkits	Molecular Field-Based Design	Molecular Docking & Virtual Screening
Founded	1990	1997	2002	1998
Headquarters	New York, USA	Santa Fe, NM, USA	Cambridge, UK	Sankt Augustin, Germany
Core Technology	FEP+, Molecular Dynamics, AI	Molecular Design Toolkits	Flare, Molecular Field Analysis	SeeSAR, Docking Algorithms
Market Position	Public (Nasdaq: SDGR)	Acquired by Cadence 2022 ($500M)	Private	Private
Revenue Scale	$238 million (2024)	Undisclosed (Part of Cadence)	Estimated $50 million	Estimated $20 million
Customer Base	All Top 20 Pharma, 1,750 customers	Biotech/Pharma focused	European Pharma focused	Academia/Small Biotech
AI Integration	Proprietary Physics-Based AI	Limited	Enhanced with Molab.ai acquisition	Traditional CompChem focused
Cloud Support	Full Google Cloud integration	Limited	Partial cloud support	On-premise focused
Proven Success	Multiple FDA-approved drugs contributed	Tool provision focused	Strong in European market	Academia research focused

Differentiation Factor Analysis

Schrödinger's Unique Strengths:

1. Physics-Based Accuracy: Over 95% accuracy with predictions based on chemical principles

2. Complete Integrated Platform: End-to-end solution from research to development

3. Proven ROI: Revenue realization through actual drug approvals and licensing deals

4. Global Partnerships: Strategic collaborations with all Top 20 pharmaceutical companies

Competitive Advantages:

• OpenEye vs Schrödinger: OpenEye focuses on tool provision, Schrödinger is a full drug development partner

• Cresset vs Schrödinger: Cresset is Europe-centered, Schrödinger is a global leader

• BioSolveIT vs Schrödinger: BioSolveIT is academia-focused, Schrödinger has proven commercial success

Global Bio Customer Status and Korean Market Entry

Global Major Customer List

Top 20 Pharmaceutical Companies (All Customers):

United States:

• Pfizer Inc.

• Johnson & Johnson

• Bristol Myers Squibb

• Merck & Co.

• AbbVie

• Gilead Sciences

• Amgen

• Eli Lilly

Europe:

• Novartis (Switzerland)

• Roche/Genentech (Switzerland)

• Sanofi (France)

• AstraZeneca (UK)

• Bayer (Germany)

• GSK (UK)

Asia-Pacific:

• Takeda (Japan)

• Otsuka (Japan)

• Daiichi Sankyo (Japan)

Biotech and Startups:

• Nimbus Therapeutics (Co-founded)

• Morphic Therapeutics (Acquired by Lilly)

• Agios Pharmaceuticals

• Multiple VC-funded biotechs

Korean Bio Drug Development Market Status

Current Drug Development Status of Major Korean Bio Companies:

Samsung Bioepis:

• World's 2nd largest in biosimilars (Based on FDA/EMA approvals)

• Completed spin-off from Samsung Biologics in 2025

• Success in Herceptin, Humira, Enbrel biosimilars

• Schrödinger Application Potential: Computer-aided design for next-generation biobetter development

Celltrion:

• World's 1st in biosimilar revenue

• Global success with Remsima, Truxima

• Expansion strategy for new drug development after Chairman Seo Jung-jin's return

• Schrödinger Utilization Areas: Antibody optimization and novel target discovery

Alteogen:

• Possesses biobetter technology NexP™ Fusion Technology

• Specializes in long-acting biopharmaceutical development

• Secured Herceptin SC formulation patents

• Schrödinger Application Areas: Protein structure-based long-acting design

Korean Market Entry Strategy:

Technical Needs:

• Korean bio companies' transition from biosimilars to new drug development

• Need for AI-based drug discovery platforms to secure global competitiveness

• Accurate predictive modeling for Regulatory Science response

Market Opportunities:

• Government's K-Bio policy support

• Bio fund formation and investment activation

• Increased collaboration with global pharmaceutical companies

Current AI Integration and Future Vision

Proprietary AI Development Focus

Schrödinger develops domain-specific AI models rather than integrating general-purpose LLMs, focusing on:

Current AI Technology Stack:

• Active Learning Applications: ML models trained on physics-based data

• LiveDesign Learning: Project-specific AI/ML model training and deployment

• De Novo Design Workflow: AI system for exploring billions of chemical spaces

Physics-Based Validation Philosophy

CEO Ramy Farid's famous quote:

"Physics-based experiments are the cake; ML is just the cherry on top. AI alone cannot create molecules that can get FDA approval."

This emphasizes the importance of verifiable AI through physics-based validation rather than black-box approaches.

Future Integration Opportunities

• Scientific Literature Analysis: Enhanced research through AI-assisted paper analysis

• Natural Language Interface: Intuitive molecular design requirement input

• Experimental Planning: AI-assisted protocol optimization

Jensen Huang's Advice and Future Vision

NVIDIA CEO Jensen Huang advised Schrödinger to "think bigger". This means leveraging AI's potential more actively.

Future Development Directions:

1. Multimodal AI Integration: AI systems integrating molecular structures, experimental data, and literature information

2. Autonomous Drug Discovery: AI automation from hypothesis generation to experimental design

3. Personalized Medicine: Patient-specific drug design considering individual genetic information

Real Success Cases and ROI Analysis

Proven Drug Development Achievements

1. Nimbus Therapeutics Co-founding Success

ACC Inhibitor Program (Firsocostat):

• Designed with Schrödinger's FEP+ technology

• Licensed to Gilead Sciences for $1.2 billion

• Schrödinger Revenue: $46 million (2016-2017)

• Current Status: Progressing as NASH treatment in Phase 2

TYK2 Inhibitor (TAK-279):

• Autoimmune disease treatment

• Acquired by Takeda for $1.1 billion

• Schrödinger Revenue: $111 million

• Current Status: Positive Phase 2b results in psoriasis

2. Bristol Myers Squibb Large-Scale Collaboration

Contract Scale: Total $2.7 billion (Industry's largest)

• Upfront: $55 million

• Milestones: Up to $2.7 billion

• Royalties: Linked to commercialized product sales

Ongoing Programs:

• Oncology: HIF-2α, SOS1-KRAS targets

• Neurology: Undisclosed CNS targets

• Immunology: Multiple inflammatory disease targets

Cost Reduction Effect Analysis

Traditional Drug Discovery vs Schrödinger Platform

Stage	Traditional Method	Schrödinger Platform	Savings
Hit Discovery	6-12 months, millions of dollars	1-2 months, hundreds of thousands	70% time savings
Lead Optimization	2-3 years, tens of millions	6-12 months, millions	80% cost savings
Candidate Selection	Total 5-6 years	Total 2-3 years	50% time reduction
Success Rate	Overall 9.6%	Improved with physics-based prediction	Reduced failure risk

ROI Calculation Case:

Georgia-Pacific Manufacturing Case:

• Traditional: Equipment failure → Manual search (30 minutes to hours)

• Schrödinger AI: "Conveyor belt strange noise" query → Instant response

  "Sensor analysis shows bearing wear. Manual section 3.2 lubrication needed. Similar case: Last month Line B same issue resolved. Recommendation: Adjust weekly inspection schedule"

  
  

Future Prospects and Market Opportunities

Global AI Drug Discovery Market Growth

Market Size:

• 2024: $28 billion

• 2028: $153 billion (IDC forecast)

• CAGR: 29.7%

Growth Drivers:

• Increased AI investment by large pharmaceutical companies

• AI platform adoption by biotech startups

• Expansion of personalized medicine

Schrödinger's Competitive Advantages

1. Unique Market Positioning

"Application-First" Strategy:

• Provides 130+ turnkey AI applications

• Rapid value realization within 3-6 months

• Industry-specific specialized solutions

Proven Business Model:

• Software licensing: $135 million annually

• Drug discovery collaboration: $45 million annually

• Royalties and milestones: Continuous revenue generation

2. Technical Differentiation

Physics-Based Accuracy:

• Reliability secured through predictions based on chemical laws

• Explainable results instead of black-box AI

• Scientific evidence for regulatory approval

Hybrid AI Approach:

• Physics-based accuracy + Machine learning speed

• Domain knowledge + Data-driven learning

• Verifiable AI systems

Korean Market Entry Opportunities

1. Connection with K-Bio Policy

Government Support Policies:

• Bio-health core technology development project

• Bio-health regulatory science support center establishment

• K-Bio vaccine and therapeutic ecosystem construction

Schrödinger Utilization Areas:

• Accurate predictive modeling for regulatory science response

• Securing global-level drug discovery capabilities

• Supporting transition from biosimilars to new drug development

2. Korean Bio Company Customized Solutions

Samsung Bioepis:

• Supporting biosimilar → biobetter transition

• FEP+ utilization for antibody structure optimization

• Next-generation biopharmaceutical design

Celltrion:

• M&A target drug candidate validation

• Novel antibody target discovery

• Existing pipeline optimization

Alteogen:

• Long-acting formulation design optimization

• Combination of NexP™ technology with Schrödinger platform

• SC formulation development acceleration

Final Recommendations: Strategic Choice for Korean Bio Industry Leap

Urgency for Competitiveness Securing

For Korean bio companies to leap into new drug development based on biosimilar success, global-level AI-based drug discovery capabilities are essential. Adopting Schrödinger platform means not just tool utilization, but transition to next-generation drug discovery paradigm.

Opportunities Korean Companies Should Not Miss:

1. First Mover Advantage: Leading Schrödinger-based AI drug discovery in Asian market

2. Global Partnerships: Expanded opportunities for direct collaboration with Top 20 pharmaceutical companies

3. Talent Acquisition: Nurturing world-class computational chemistry/AI experts

4. Regulatory Competitiveness: Improving regulatory approval probability with AI-based predictive models

Specific Action Plan for Implementation

Immediately Executable Measures:

1. Pilot Project Selection: Determining Schrödinger application priorities among existing pipelines

2. Expert Recruitment: Securing computational chemistry/structural biology PhD personnel

3. Collaboration with Schrödinger Korea Office: Leveraging local support through Korean operations

4. Education Programs: Schrödinger platform training for existing research staff

Mid-to-Long Term Strategy:

1. R&D Center Establishment: Creating dedicated AI drug discovery organization

2. Cloud Infrastructure: Building scalable computing environment based on Google Cloud

3. Global Network: Participating in collaboration networks with Schrödinger customers

4. IP Portfolio: Securing Schrödinger-based drug discovery patents

Conclusion: New Era Drug Discovery Standard

Schrödinger presents new standards for drug discovery by combining 30 years of scientific expertise with cutting-edge AI technology. Predictive modeling based on physics-based accuracy overcomes the limitations of existing AI tools and enables the design of drugs that can actually receive FDA approval.

For Korean bio companies, Schrödinger is not just a software tool, but a strategic partner for securing global competitiveness. Major companies like Samsung Bioepis, Celltrion, and Alteogen can utilize this opportunity to connect biosimilar success to drug discovery leap.

Future drug discovery will be realized through perfect harmony of physics and AI, and Schrödinger will be the core platform supporting Korean bio industry's global leadership at its center.

ⓒ 2025 The intellectual property rights of this report belong to the author and respective companies.