Can a 3D image network predict how mutations change protein stability?
Training a 3D convolutional network on both each mutation and its reverse removed the usual bias toward predicting destabilisation, and removing related proteins from training showed earlier methods' scores were likely inflated.
Source
Predicting changes in protein thermodynamic stability upon point mutation with deep 3D convolutional neural networks
Study at a glance
- Design
- Computational / modelling — Ensemble of ten 3D CNNs on voxelised wild-type and mutant structures, trained on homology-pruned data augmented with reverse mutations; compared with seventeen predictors on a balanced blind test set, plus ablations and case studies.
- N
- N=684 · 684 test mutations in S_sym (342 direct plus their reverses); training used 3,488 points (1,744 direct mutations plus reverses); case studies of 42 p53 and 134 myoglobin mutations; ClinVar variants for an application analysis.
- Population
- Experimentally measured single-point protein mutations and their structures
- Outcome
- Pearson correlation and RMSE between predicted and measured stability change (ΔΔG), and anti-symmetry bias between direct and reverse mutations
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The authors built ThermoNet, which voxelises the 3D neighbourhood of a mutation site in both the original and a computer-modelled mutant structure, encodes atom properties as channels, and feeds this to an ensemble of ten 3D CNNs to predict the stability change. They pruned training proteins that were homologous to test proteins and doubled the data by adding each mutation's reverse, whose stability change is the negative. They tested on a balanced benchmark of direct and reverse mutations against seventeen other predictors, ran an ablation trained on direct mutations only, and applied the model to p53, myoglobin and ClinVar variants.
What they found
ThermoNet reached a correlation of about 0.47 with measured values for both direct and reverse mutations, with almost perfect anti-symmetry between the two (correlation of −0.96). A version trained only on direct mutations did similarly on direct mutations but essentially failed on reverse ones (r about −0.06). A version trained without homology pruning scored better on the same test (reverse r 0.59 vs 0.48), showing leakage inflates results. On ClinVar, ThermoNet placed 80.2% of benign variants in a near-neutral stability zone, versus 39.7% for FoldX.
The limits
What it doesn't show
Accuracy is moderate (correlations near 0.5), and after homology pruning two other methods, DDGun and DDGun3D, outperformed ThermoNet; the authors could not retrain competitors on the pruned data, so comparisons are not like-for-like. The method needs an experimental or modelled 3D structure and relies on Rosetta-generated mutant models rather than real mutant structures. The ClinVar analysis checks consistency with a theoretical model of protein evolution, not against measured stability values. Training data are small for a 3D CNN with over a hundred thousand parameters.
Key terms
- ΔΔG
- The change in a protein's folding free energy caused by a mutation; the sign says whether the mutation stabilises or destabilises the protein.
- Anti-symmetry
- The physical rule that the reverse mutation (mutant back to original) has exactly the opposite ΔΔG of the forward one.
- Data augmentation
- Creating extra training examples from existing ones, here by adding each mutation's reverse with the negated label.
- Homology leakage
- When training and test sets contain evolutionarily related proteins, so the test no longer measures performance on truly new proteins.
- 3D convolutional neural network
- A network whose filters slide through a 3D grid, detecting local spatial patterns such as atom arrangements.
Flashcards
0 of 11 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
How did ThermoNet reduce bias toward destabilising predictions?
Common questions
Why were earlier predictors biased toward destabilising?
Their training data were dominated by destabilising mutations, so models learned to lean that way; they did well on direct mutations but badly on reverse ones.
Why did ThermoNet* score higher if ThermoNet is the better method?
ThermoNet* was trained with proteins related to the test proteins left in, so part of its higher score reflects leakage rather than real generalisation.
Where do the mutant structures come from?
They are modelled from the original structure with the Rosetta software, since experimental mutant structures are rare.
More on Convolutional networks