ESMIFDesign
ESMDesign uses ESM-IF1 for the design of fragments within multichain proteins
Install / Use
/learn @LBC-LNBio/ESMIFDesignREADME
ESMIFDesign
This repository focuses on designing T-cell receptors (TCRs) using the ESM-IF1 deep learning method.
The ESM-IF1 inverse folding method is built for predicting protein sequences from their backbone atom coordinates. Here, we use the ESM-IF1 model (esm_if1_gvp4_t16_142M_UR50 - fair-esm v2.0.1) to design part of the TCRs sequences, considering their peptide-major histocompatibility complex (pMHC) complex. Positions outside the specified criteria were held constant. The designed positions were changed to <mask> tokens, and all amino acid substitutions, including cysteine, were allowed. Given the multi-chain structure, a padding of 10 <pad> tokens was used to separate the chains.
Dependencies
To install the dependencies, run:
pip install -r requirements.txt
Usage
The configuration file is located at config.json, where it specifies, for each structure, the residues and chains to be designed. The file is organized as follows:
{
"6zkw": ["110D","111D","112D","134D","135D","113E","114E","133E"],
...,
"8shi": ["109D","110D","111D","112D","113D","114D","135D","110E","111E","112E","113E","134E","135E"]
}
To design TCR sequences, run:
python run.py
Testing
We tested some conditions to check the performance of the model.
- Testing on
6ZKWstructure:6ZKW.pdbrepresents a TCR-pMHC complex.6ZKW_DE.pdbincludes only the TCR (D and E chains).6ZKW_contact_to_gly.pdbis a TCR-pMHC complex with mutations on the TCR residues in contact with the pMHC, changing them to glycine.6ZKW_all_gly.pdbis a TCR-pMHC complex with mutations on all residues, changing them to glycine.
- Testing on pMHC interference on protein sequence design (
tests/pMHC1.config) - Testing temperature in the interval (
tests/temperature.config): [1e-6, 1e-5, 1e-4, 1e-3, 1e-2, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 2, 5] - Testing number of samples in the interval (
tests/sampling.config): [5, 10, 25, 50, 100, 250, 500] - Testing design approaches:
CDR3 interface(tests/CDR3_interface.config): restricting the design to CDR3 (α and β TCR chains) within a proximity of 5 Å to either the peptide or MHC;CDR3(tests/CDR3.config): designing the entire CDR3 (α and β TCR chains).CDRs interface(tests/CDRs_interface.config): designing CDR1, CDR2, or CDR3 positions (α and β TCR chains) within a 5 Å distance to the peptide or MHC;
To run the tests, navigate to the tests directory and execute:
cd tests
python testing.py
Related Skills
diffs
341.2kUse the diffs tool to produce real, shareable diffs (viewer URL, file artifact, or both) instead of manual edit summaries.
clearshot
Structured screenshot analysis for UI implementation and critique. Analyzes every UI screenshot with a 5×5 spatial grid, full element inventory, and design system extraction — facts and taste together, every time. Escalates to full implementation blueprint when building. Trigger on any digital interface image file (png, jpg, gif, webp — websites, apps, dashboards, mockups, wireframes) or commands like 'analyse this screenshot,' 'rebuild this,' 'match this design,' 'clone this.' Skip for non-UI images (photos, memes, charts) unless the user explicitly wants to build a UI from them. Does NOT trigger on HTML source code, CSS, SVGs, or any code pasted as text.
openpencil
1.9kThe world's first open-source AI-native vector design tool and the first to feature concurrent Agent Teams. Design-as-Code. Turn prompts into UI directly on the live canvas. A modern alternative to Pencil.
HappyColorBlend
HappyColorBlendVibe Project Guidelines Project Overview HappyColorBlendVibe is a Figma plugin for color palette generation with advanced tint/shade blending capabilities. It allows designers to
