20240029823. PEPTIDE BASED VACCINE GENERATION SYSTEM WITH DUAL PROJECTION GENERATIVE ADVERSARIAL NETWORKS simplified abstract (NEC Laboratories America, Inc.)
Contents
PEPTIDE BASED VACCINE GENERATION SYSTEM WITH DUAL PROJECTION GENERATIVE ADVERSARIAL NETWORKS
Organization Name
NEC Laboratories America, Inc.
Inventor(s)
Renqiang Min of Princeton NJ (US)
Hans Peter Graf of South Amboy NJ (US)
PEPTIDE BASED VACCINE GENERATION SYSTEM WITH DUAL PROJECTION GENERATIVE ADVERSARIAL NETWORKS - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240029823 titled 'PEPTIDE BASED VACCINE GENERATION SYSTEM WITH DUAL PROJECTION GENERATIVE ADVERSARIAL NETWORKS
Simplified Explanation
The abstract describes a computer-implemented method for generating new binding peptides to major histocompatibility complex (MHC) proteins. The method involves training a generative adversarial network (GAN) on a set of binding peptide sequences and non-binding peptide sequences. The GAN training objective updates the discriminator to distinguish between generated and sampled binding peptide sequences, while the generator is updated to fool the discriminator. The training also optimizes the GAN training objective by learning two projection vectors for a binding class using cross-entropy losses.
- The method uses a generative adversarial network (GAN) to generate new binding peptides to MHC proteins.
- The GAN consists of a generator and a discriminator.
- The discriminator is trained to distinguish between real binding peptide sequences and generated/fake peptide sequences.
- The generator is trained to generate peptide sequences that can fool the discriminator.
- The training includes optimizing the GAN training objective using two projection vectors and cross-entropy losses.
- The first loss discriminates between binding and non-binding peptide sequences in the training data.
- The second loss discriminates between generated binding peptide sequences and non-binding peptide sequences in the training data.
Potential applications of this technology:
- Drug discovery: The generated binding peptides can be used to design new drugs that target specific MHC proteins.
- Vaccine development: The generated binding peptides can be used to develop vaccines that stimulate immune responses against specific pathogens.
- Personalized medicine: The generated binding peptides can be used to design personalized therapies based on an individual's MHC profile.
Problems solved by this technology:
- Limited peptide diversity: Traditional methods for generating binding peptides have limited diversity, but this technology can generate a wide range of new binding peptides.
- Time and cost efficiency: The computer-implemented method can generate binding peptides more quickly and at a lower cost compared to traditional experimental methods.
Benefits of this technology:
- Enhanced drug discovery: The ability to generate new binding peptides can lead to the discovery of novel drugs with improved efficacy and specificity.
- Customized therapies: The generated binding peptides can be tailored to an individual's MHC profile, enabling personalized treatment options.
- Accelerated research: The computer-based approach allows for rapid generation and screening of binding peptides, accelerating the research and development process.
Original Abstract Submitted
a computer-implemented method is provided for generating new binding peptides to major histocompatibility complex (mhc) proteins. the method includes training, by a processor device, a generative adversarial network gan having a generator and a discriminator only on a set of binding peptide sequences given training data comprising the set of binding peptide sequences and a set of non-binding peptide sequences. a gan training objective includes the discriminator being iteratively updated to distinguish generated peptide sequences from sampled binding peptide sequences as fake or real and the generator being iteratively updated to fool the discriminator. the training includes optimizing the gan training objective while learning two projection vectors for a binding class with two cross-entropy losses. a first loss discriminating binding peptide sequences in the training data from non-binding peptide sequences in the training data. a second loss discriminating generated binding peptide sequences from non-binding peptide sequences in the training data.
