1 of 13

PeptGPT

Accelerating Protein Engineering with GPT-4

*A novel way to design proteins*

-Cal Hacks 2023 Judge

2 of 13

-Problem-

Biotech researchers waste valuable time

designing proteins for experiments

3 of 13

-Our Solution-

Use ChatGPT-4 and

Large Language Models (LLMs)

to design proteins

4 of 13

A More Efficient Design Process

Identify Desired Function

Literature Search

Candidate Protein Families

Protein Sequence Generation

Test Protein Folding

Identify Desired Function

Single Fusion Protein Sequence

Protein Folding in ESM2

Usual

timeline:

PeptGPT

timeline:

5 of 13

We make in minutes

what researchers made in weeks

Save Time with AI

6 of 13

Our Design

User enters keyword into GPT-4 to produce Protein Family (PFAM) value

Interpro searches databases for specific protein sequence by PFAM value

Probability statistics assembles each part of a large protein sequence

ESM-2 folds protein sequence

ESM-2 Meta

7 of 13

Our Competitors

overproduce protein sequences,

While we prioritize design by function

8 of 13

Competition

Focus on Protein Design

Cheap

Accurately Generate Proteins

Fast

Easy to Use

AI Gene Circuit Design

AI Protein Design

PeptGPT

9 of 13

Protein Engineering’s Fiscal Value

Market Value (2023)

$3 Billion

Expected Market Value (2033)

$10 Billion

Current Compound Annual Growth (CCAG) rate

13.6%

10 of 13

Project Timeline

Y1: Funding & Networking

Y0: Develop Product

Y2: Market Product

Y4: Company Expansion

Y5: Company Stabilization

Y3: Company Expansion

%

11 of 13

Board of Founders

Nathaniel Sheps,

BS Chemical Engineering Johns Hopkins University (WIP)

Alexander Chow,

BA Computer Science

UC Berkeley (WIP)

Sreenivas Eadara,

MD-PhD Biomedical Engineering Northwestern University (WIP)

Rajashekar Vennavelli,

MS Computer Science Santa Clara University (WIP)

12 of 13

We Need Your Help

With Batch-17’s networking and financial aid,

we can dominate the protein engineering market

13 of 13

PeptGPT