公开(公告)号：WO2022169681A1

公开(公告)日：2022-08-11

申请号：PCT/US2022/014211

申请日：2022-01-28

Applicant: NEC LABORATORIES AMERICA, INC.

Inventor： MIN, Renqiang ,  GRAF, Hans Peter ,  HAN, Ligong

IPC: G06N3/08

Abstract: A method for learning disentangled representations of videos is presented. The method includes feeding (1001) each frame of video data into an encoder to produce a sequence of visual features, passing (1003) the sequence of visual features through a deep convolutional network to obtain a posterior of a dynamic latent variable and a posterior of a static latent variable, sampling (1005) static and dynamic representations from the posterior of the static latent variable and the posterior of the dynamic latent variable, respectively, concatenating (1007) the static and dynamic representations to be fed into a decoder to generate reconstructed sequences, and applying (1009) three regularizes to the dynamic and static latent variables to trigger representation disentanglement. To facilitate the disentangled sequential representation learning, orthogonal factorization in generative adversarial network (GAN) latent space is leveraged to pre-train a generator as a decoder in the method.

公开(公告)号：WO2022216584A1

公开(公告)日：2022-10-13

申请号：PCT/US2022/023264

申请日：2022-04-04

Applicant: NEC LABORATORIES AMERICA, INC.

Inventor： MIN, Renqiang ,  GRAF, Peter Hans ,  HAN, Ligong

IPC: G16B40/00 ,  G16B15/30 ,  G06N3/04 ,  G06N3/08

Abstract: A method is provided for generating new binding peptides to MHC proteins includes training (430), 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 including 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 (440) 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.

公开(公告)号：WO2022216592A1

公开(公告)日：2022-10-13

申请号：PCT/US2022/023281

申请日：2022-04-04

Applicant: NEC LABORATORIES AMERICA, INC.

Inventor： MIN, Renqiang ,  GRAF, Hans Peter ,  HAN, Ligong

IPC: G16B40/00 ,  G16B30/00 ,  G16B5/00 ,  G16B15/30 ,  G01N33/68

Abstract: Methods and systems for training a machine learning model include embedding (304) a state, including a peptide sequence and a protein, as a vector. An action, including a modification to an amino acid in the peptide sequence, is predicted (306) using a presentation score of the peptide sequence by the protein as a reward. A mutation policy model is trained (308), using the state and the reward, to generate modifications that increase the presentation score.

公开(公告)号：WO2022216591A1

公开(公告)日：2022-10-13

申请号：PCT/US2022/023280

申请日：2022-04-04

Applicant: NEC LABORATORIES AMERICA, INC.

Inventor： MIN, Renqiang ,  GRAF, Hans Peter ,  HAN, Ligong

IPC: G16B30/00 ,  G16B40/00 ,  G06N3/08 ,  G06N3/04

Abstract: Methods and systems for training a model include encoding (203) training peptide sequences using an encoder model. A new peptide sequence is generated (202) using a generator model. The encoder model, the generator model, and the discriminator model are trained (206) to cause the generator model to generate new peptides that the discriminator mistakes for the training peptide sequences, including learning projection vectors with respective cross-entropy losses for binding sequences and non-binding sequences.