TIMED-Design: Flexible and Accessible Protein Sequence Design with Convolutional Neural Networks

Models for the paper "**TIMED-Design: Flexible and Accessible Protein Sequence Design with Convolutional Neural Networks**"

# Performance Comparison

For detailed performance comparisons, please see the paper.

## Macro Recall
Macro-Recall is accuracy averaged per residue - resistant to class imbalance.

## RMSD_100
We sampled 10% of the PDBench dataset and ran it through AlphaFold2 + Amber relaxation. RMSD_100 is a normalised version of RMSD.

## Isoelectric Point Mean Absolute Error (MAE)
Difference between the isoelectric point of the original sequence and the predicted sequence.

## Charge Mean Absolute Error (MAE)
Difference between the overall charge of the original sequence and the predicted sequence.

# Training

All models were trained using the culled PDB set from PISCES `cullpdb_pc90_res3.0_R1.0_d200702_chains40583`containing over 35K non-redundant protein structures (40K+ chains), with resolutions up to 3.0 Å.

## CNN Models
We reimplemented all of the CNN models in the literature as they were all closed-source. The dataset for CNN models was created using [aposteriori](https://github.com/wells-wood-research/aposteriori/) using the following command:

```
make-frame-dataset /scratch/datasets/biounit/ -d benchmarking_set.csv -e .pdb1.gz --voxels-per-side 21 --frame-edge-length 21 -g True -p 35 -n benchmark_set -v -r -z -cb True -ae CNOCBCA  --compression_gzip True -o /scratch/timed_dataset/
```

For Charge and Polar models we used the codecs (`-ae`) equivalent to `CNOCBCAQ` and  `CNOCBCAP`, respectively. 

## GNN Models
The code for training ProteinMPNN with custom training sets is not available. We recreated the steps given to us by the authors and published them here: https://github.com/wells-wood-research/ProteinMPNN_custom_training/tree/main

## What's Changed
* Add output_dir as functionality by @universvm in https://github.com/wells-wood-research/timed-design/pull/66
* Fix .fasta files output by @LunaPrau in https://github.com/wells-wood-research/timed-design/pull/68
* Simplify Install  by @universvm in https://github.com/wells-wood-research/timed-design/pull/62
* Fix security vulnerabilities by @universvm in https://github.com/wells-wood-research/timed-design/pull/69
* Hide streamlit warnings by @universvm in https://github.com/wells-wood-research/timed-design/pull/71
* Fix docker by @universvm in https://github.com/wells-wood-research/timed-design/pull/70
* Hide charge and polar until #64 is merged by @universvm in https://github.com/wells-wood-research/timed-design/pull/73
* Add page title. by @ChrisWellsWood in https://github.com/wells-wood-research/timed-design/pull/75

## New Contributors
* @LunaPrau made their first contribution in https://github.com/wells-wood-research/timed-design/pull/68
* @ChrisWellsWood made their first contribution in https://github.com/wells-wood-research/timed-design/pull/75

**Full Changelog**: https://github.com/wells-wood-research/timed-design/compare/modelspublication...publication_01_2024