Mutation-induced biophysical destabilization as a key contributor to cancer-driving potential in the human structural protein interactome
Su, T.-Y.; Xia, Y.
Show abstract
A systems-level investigation of mutation-induced perturbations in the human structural protein interactome, the network of structurally resolved protein-protein interactions in human cells, provides mechanistic insight into the complexity underlying oncogenesis. Mutations can destabilize protein folding or specific protein-protein interactions, resulting in loss-of-function effects within the interactome. Although such network-level loss-of-function consequences may contribute to oncogenesis, previous studies have largely examined either folding or binding destabilization in isolation. Here, we performed structural and free-energy calculations to assess the impact of interactome-wide cancer-associated missense mutations on protein-folding stability and protein-binding stability. We assessed the cancer-driving potential of destabilizing mutations using a "fold difference" metric, defined as the ratio between the fractions of destabilizing mutations in datasets of cancer-associated and non-pathogenic mutations. We observed a strong positive correlation between biophysical destabilization and cancer-driving potential, with folding-destabilizing ("quasi-null") and binding-destabilizing ("edgetic") mutations within cancer-driving genes causing stronger structural and functional effects than those across the entire cancer genome. Our findings align with the expectation that cancer-driving genes are enriched in driver mutations and suggest that biophysical destabilization is a key contributor to cancer-driving potential. Overall, our study provides a biophysical perspective on the loss-of-function implications of destabilizing mutations across the interactome, enhancing comprehension of the intricate processes driving oncogenesis.
Matching journals
The top 9 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Rhapsody: Pathogenicity prediction of human missense variants based on protein sequence, structure and dynamics 95%
- Transfer learning via multi-scale convolutional neural layers for human-virus protein-protein interaction prediction 92%
- PathExt: a general framework for path-based mining of omics-integrated biological networks 92%
Similar papers in this journal
Similar papers in this journal
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.