700,000 deaths/year from antibiotic resistance — and rising

Identify druggable targets in
antibiotic resistance proteins

Problem

Druggability analysis and literature mining are fragmented across incompatible tools.

Solution

ResistAI integrates AlphaFold, ESM-2 embeddings, fpocket, and RAG into one automated pipeline.

Outcome

Ranked druggable pockets + AI-cited literature in seconds — for any WHO priority pathogen.

AlphaFold predicts the structure. UniProt holds the sequence.ResistAI tells you whether it's druggable — and runs the analysis live for any protein.

Loading...
drag to rotate · scroll to zoom
2,433
Proteins analysed
1,198
High druggability
2,508
PubMed articles
1.000
Best score

Why ResistAI?

ResistAI does this automatically — saving hours of fragmented manual work.

Without ResistAI

  • Go to AlphaFold, download the structure manually
  • Install & run fpocket locally, parse raw output files
  • Search PubMed separately, read through dozens of abstracts
  • Manually integrate structure, pockets, and literature findings
  • No way to find evolutionarily similar resistance targets

With ResistAI

  • Druggability scoring — AlphaFold gives structure; ResistAI ranks 2,433 pockets automatically
  • AI literature summary — Llama 3.3 synthesises PubMed findings with PMID citations in seconds
  • ESM-2 similarity search — find evolutionarily related resistance targets instantly
  • On-demand druggability — enter any UniProt ID and fpocket scores it live; AlphaFold gives a structure, ResistAI gives the druggable pockets
  • One platform — structure + pockets + embeddings + literature, fully integrated

How it works

From protein identifier to ranked druggable targets — fully automated, reproducible, and AI-native.

Built with

Nextflow DSL2AlphaFold DBfpocketESM-2ChromaDBLlama 3.3 70BFastAPINext.jsDockerSlurmSupabase

Platform features

Start your research today

Free access to 2,433 analysed resistance proteins, 2,508 indexed articles, and AI-powered literature synthesis.

Create free account →