Research section Research projects and references

Research: NIPAH-VIRUS-DECOY-PROTEIN-DYNAMICS Folding Project #13021

Project #13021 overview

Project Summary AI Beta

Nipah virus is deadly and spreads easily. This project uses computer models to design decoy proteins that block the virus from infecting cells. Simulations help us understand how these designed proteins work better than the natural ones.
Automated summary; simplified and may not be fully accurate.

Project team

Manager(s)
Lin Zhu
Institution
the University of Illinois Urbana-Champaign

Work unit

Atoms
60,000
Core
0x24
Status
Beta
Source material

Official Project Description

Nipah virus (NiV) is a highly dangerous pathogen first identified in the late 1990s that poses a significant threat to global health.

Naturally carried by fruit bats, the virus can spread to humans directly, through intermediate hosts like pigs, or via person-to-person contact.

Infections are often fatal, with mortality rates between 40% and 80%, typically resulting from severe respiratory issues or acute brain inflammation (encephalitis).

The World Health Organization has designated it a priority disease, signaling an urgent global need for accelerated research into effective medical countermeasures to prevent a potential pandemic. This project uses a data-driven approach to engineer "decoy" proteins that neutralize the Nipah Virus (NiV) by mimicking EFNB2, the virus's natural receptor for human cells.

While our machine learning models successfully identify effective variants 70% of the time, the physical reasons behind their success remain unclear.

To bridge this gap, we use molecular dynamics simulations to examine these engineered proteins at the atomic level.

By comparing their conformational dynamics with those of the wild-type protein, we investigate how engineered mutations reshape the conformational landscape and alter structural flexibility.

These simulations allow us to identify mutation-induced conformational changes and characterize the molecular interactions that stabilize functionally relevant states, providing mechanistic insight into the enhanced performance of the engineered variants.

Performance data

Hardware Performance for Project 13021

Compare community-sampled Folding@Home output for the GPUs and CPUs processing this project.

Data as of Thursday, 03 September 2026 18:31:30

GPU PPD Averages

Rank
Project
Model Name
Folding@Home Identifier
Make
Brand
GPU
Model
PPD
Average
Points WU
Average
WUs Day
Average
WU Time
Average
1 Intel Arc B580 Graphics
Battlemage G21 [Intel Arc B580 Graphics]
Intel Battlemage G21 3,353,135 60,313 55.60 0 hrs 26 mins
2 GeForce RTX 2060
TU106 [Geforce RTX 2060]
Nvidia TU106 2,134,772 259,976 8.21 2 hrs 55 mins
3 GeForce GTX 1050 LP
GP107 [GeForce GTX 1050 LP] 1862
Nvidia GP107 383,424 60,313 6.36 3 hrs 47 mins
4 GeForce GTX 1070 Mobile
GP104BM [GeForce GTX 1070 Mobile] 6463
Nvidia GP104BM 199,627 60,313 3.31 7 hrs 15 mins