Mewtate - Impact of viral protein sequence variants on protein domains and interfaces

Project Contributors: Houcemeddine Othman, Sachendra Kumar, Sherlyn Jemimah, Philippe Youkharibache, Xavier Watkins

Motivation

Tool to evaluate impact of a mutation

• Come up with a simple SNPs variants mutations structural analysis

• Application to ACE2 polymorphism vs. SARS-CoV-2 binding

  • After the identification of an interaction network between 2 sets of residues (general/flexible case) 2 cases can be considered
    • A mutation disrupts the network << understand the disruptions using a change of residue side chain corresponding to a mutation (for example P>F) sampling rotamers and corresponding clashes/Hbonds etc formed/broken (with local optimization possible)
    • Use coevolving mutations to evolve an interface given a change from one species to another or a species polymorphism on one protein, adapting its partner
      • Application between bat/pangolin.civet/human in the couple ACE2/SARS-CoV

What does Mewtate do?

Mewtate is a tool to predict the effect of single amino acid substitution (point mutation) on 3D protein structures. It also allows user to identify mutation hotspot regions by calculating statistically significant cluster of mutations on a 3D protein structure.

Mewtate Workflow

Mewatate assess structural changes such as disulfide breakage, seconday structure change, and many more structural features upon point mutation in a protein structure. This tools also calculate the change in free energy due to point mutation and its effect on interaction between protein complexes using FoldX. In addition, user can identify mutation hotspot region by calculating internal distribution of mutations for structurally important region compared with other regions of same proteins using statistical approach. Mewtate allows user to prioritize mutations for experimental validation.

Application architecture

Backend

The application backend is written in Python and uses FoldX and DSSP to make its calculations. It is accessed through a Flask server and returns a json response once the calculations are finished.

Client

The mutate client is a React application which connects to the backend to submit user input. When a user selects a SARS-CoV-2 protein from the dropdown menu, a request is made to the PDBe API to retrieve the list of available structures. These then appear in another dropdown, and the user can then select the structure and chain they would like to test the mutation on. They can then input the residue number and amino-acid change.

A request is made to the backend to start the calculations (change in binding/structure energy as well as structure prediction). At the same time, a request is made to the UniProt API to retrieve the anotated features which overlap with the residue position.

The user is presented with a "variant report" which compiles this information, as well as integrating the IcN3D viewer.

How to use Mewtate?

Validation

In our preliminary studies, we have used ACE2/SARS-CoV to validate our results. We would extend our study other 3D protein structures. We will add more features to improve the functionality of mewtate.

References

1. Ittisoponpisan et al. 2019 https://doi.org/10.1016/j.jmb.2019.04.009
2. Miller max acc: Miller et al. 1987 https://doi.org/10.1016/0022-2836(87)90038-6
3. Wilke: Tien et al. 2013 https://doi.org/10.1371/journal.pone.0080635
4. Eduard et al. 2014 https://doi.org/10.1093/bioinformatics/btu499
5. FoldX : http://foldxsuite.crg.eu/