dominicp6 Goto Github PK
Name: Dom Phillips
Type: User
Bio: Google Scholar: https://scholar.google.com/citations?user=pMijm0cAAAAJ&hl=en
Name: Dom Phillips
Type: User
Bio: Google Scholar: https://scholar.google.com/citations?user=pMijm0cAAAAJ&hl=en
Comparing the relative performance of different collective variable learning algorithms using Kramer's rate analysis and enhanced sampling.
Dozens of vaccines protecting against SARS-CoV-2 have now been approved for public use, yet there remains a high risk that the virus evolves to escape vaccine protection. This motivates the need for a new generation of vaccines that can protect against a wider gamut of a virus’s evolutionary accessible states, not just the currently circulating strains. Computational methods such as sequence generative models can play a critical role in mapping out this state space. In particular, they can be used to screen thousands of examples of viral proteins that might pose a high risk of vaccine escape. In this work, we take steps towards such a computational method by designing and evaluating a conditional Variational Autoencoder (VAE) capable of selectively generating SARS-CoV-2 spike proteins with low immune visibility. The model is trained on $65,000$ of the most common wild-type SARS-CoV-2 sequences and uses NetMHCpan to estimate levels of exposure to human T cell immunity. The model's generated sequences are compared with those derived from two simpler generative models; a random-mutator and an 11-gram language model. We discover that although all three models are able to generate stable, structurally valid sequences, only the VAE model can generate low immunogenicity sequences sampled from a distribution that interpolates smoothly along the principal variance directions of natural sequences.
All markdown data for my Obsidian notes.
Implements parallelised classes for Overdamped Langevin Dynamics, Underdamped Langevin Dyanamics and Gaussian Drift Diffusion Dynamics in arbitary dimensions.
Langevin integrators implemented in Julia.
Visualisation and evaluation of a VAE fully generative models for Spike protein sequences for SARS-CoV-2.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.