FIDENIT
BIOLOGY · Building pipeline

Protein Structure Prediction

Structures of hard-to-drug protein targets

PROBLEM

Could we see the structure of proteins that drugs have not reached?

The AlphaFold Database openly shares more than 200 million predicted protein structures. Building on that, we aim to narrow down the structure and likely binding sites of targets no drug has reached yet, so researchers can decide where to experiment first. If we publish our code and evaluation method, others can check the work and build on it.

HOW IT WORKS
01 · INPUT

Connect

Amino acid sequence

02 · ENGINE

Refine · Calculate

We aim to connect and clean the sequence data, then have agents compute structures and check each other's work.

03 · OUTPUT

Result

3D structure and binding-site candidates, with the evidence behind them

GOALS
01

Structure from sequence

02

Narrowing binding sites

03

Reviewing candidate molecules

04

Prioritizing experiments

SPEC
FieldBiology
StatusBuilding pipeline
InputAmino acid sequence
Output3D structure · binding-site candidates
DeliveryJoint research
MORE IN BIOLOGY

Let's talk about
Protein Structure Prediction

business@fidenit.com