ANTIGEN PREDICTIONS FOR INFECTIOUS DISEASE-DERIVED EPITOPES
Disclosed herein is a system and methods for determining the alleles, antigens, and infectious disease-based vaccine composition as determined on the basis of a patient's expressed HLA alleles. Additionally described herein are unique infectious disease-derived vaccines.
1 . A method for identifying one or more infectious disease-derived antigens likely to be presented by cells of a subject, the method comprising:
obtaining peptide sequences of a plurality of infectious disease-derived antigens;
obtaining sequences of one or more MHC alleles of the subject;
inputting the peptide sequences of the plurality of infectious disease-derived antigens and the sequences of one or more MHC alleles of the subject into a multi-part presentation model to generate a set of numerical likelihoods that the plurality of infectious disease-derived antigens are presented by the one or more MHC alleles expressed on surfaces of cells of the subject,
wherein a first part of the multi-part presentation model comprises a pan-allele model portion that receives, as input, peptide sequences of one or more infectious disease-derived antigens and the sequences of one or more MHC alleles of the subject, or representations thereof, and
wherein a second part of the multi-part presentation model comprises a plurality of allele-specific models that each receives, as input, the peptide sequences of the plurality of infectious disease-derived antigens, or a representation thereof; and
selecting a subset of the plurality of infectious disease-derived antigens based on the set of numerical likelihoods to generate a set of selected antigens.
2 . The method of claim 1 , wherein the multi-part presentation model comprises a plurality of parameters generated using at least 1) mass spectrometry data and 2) binding affinity data determined from a plurality of samples.
3 . The method of claim 1 , wherein the multi-part presentation model comprises a plurality of parameters generated using a training dataset comprising:
training peptide sequences, and
for one or more of the training peptide sequences, a label derived from mass spectrometry data indicating whether the training peptide sequence was presented by one or more class I MHC alleles present in a plurality of samples.
4 . The method of claim 3 , wherein the training peptide sequences are identified through mass spectrometry on isolated peptides eluted from MHC alleles present in the plurality of samples.
5 . The method of claim 3 , wherein the multi-part presentation model comprises a plurality of parameters generated using a training dataset comprising:
for one or more of the training peptide sequences, a label derived from binding affinity data indicating whether the training peptide sequence was bound with one or more class I MHC alleles present in a plurality of samples.
6 - 8 . (canceled)
9 . The method of claim 1 , wherein the pan-allele model portion comprises a neural network.
10 . The method of claim 9 , wherein a first set of layers of the neural network of the pan-allele model portion performs a dimensional reduction of the sequences of one or more MHC alleles of the subject.
11 . The method of claim 9 , wherein a second set of layers of the neural network of the pan-allele model portion receives, as input, a representation of the peptide sequences of the plurality of infectious disease-derived antigens and a dimensionally reduced representation of the sequences of one or more MHC alleles of the subject.
12 . The method of claim 11 , wherein the representation of the peptide sequences of the plurality of infectious disease-derived antigens is generated by encoding the peptide sequences via a one-hot encoding scheme.
13 . The method of claim 11 , wherein the second set of layers of the neural network models interactions between the peptide sequences of the plurality of infectious disease-derived antigens and the sequences of one or more MHC alleles of the subject.
14 . The method of claim 1 , wherein one or more of the allele-specific models comprise a neural network.
15 . The method of claim 14 , wherein the neural network of the allele-specific network receives, as input, a representation of the peptide sequences of the plurality of infectious disease-derived antigens, and outputs per-allele presentation likelihoods for an allele.
16 . The method of claim 15 , wherein the representation of the peptide sequences of the plurality of infectious disease-derived antigens is generated by encoding the peptide sequences via a one-hot encoding scheme.
17 - 21 . (canceled)
22 . The method of claim 1 , wherein the cells of the subject comprise cells infected with one of a pathogen, virus, bacteria, fungus, or a parasite.
23 . The method of claim 1 , wherein infectious disease-derived antigens originate from one of a pathogen, virus, bacteria, fungus, or a parasite.
24 . The method of claim 1 , wherein infectious disease-derived antigens originate from an infectious disease organism selected from the group consisting of: severe acute respiratory syndrome-related coronavirus (SARS), severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), Ebola, HIV, Hepatitis B virus (HBV), influenza, Hepatitis C virus (HCV), Human papillomavirus (HPV), Cytomegalovirus (CMV), Chikungunya virus, Respiratory syncytial virus (RSV), Dengue virus, a orthymyxoviridae family virus, tuberculosis, pancorona, herpes simplex virus infection (HSV), flu, metapneumovirus (MPV), and Parainfluenza Viruses (PIVs).
25 . A method of treating a subject for an infectious disease comprising performing claim 1 , and further comprising obtaining a vaccine comprising the set of selected antigens, and administering the vaccine to the subject.
26 - 27 . (canceled)
28 . A method of manufacturing a vaccine, comprising performing claim 1 , and further comprising producing or having produced a vaccine comprising the set of selected antigens.
29 . A vaccine comprising a set of selected antigens selected by performing the method claim 1 .
30 - 33 . (canceled)
34 . A non-transitory computer readable medium comprising instructions that, when executed by a processor, cause the processor to:
obtain peptide sequences of a plurality of infectious disease-derived antigens;
obtain sequences of one or more MHC alleles of the subject;
input the peptide sequences of the plurality of infectious disease-derived antigens and the sequences of one or more MHC alleles of the subject into a multi-part presentation model to generate a set of numerical likelihoods that the plurality of infectious disease-derived antigens are presented by the one or more MHC alleles expressed on surfaces of cells of the subject,
wherein a first part of the multi-part presentation model comprises a pan-allele model portion that receives, as input, peptide sequences of one or more infectious disease-derived antigens and the sequences of one or more MHC alleles of the subject, or representations thereof, and
wherein a second part of the multi-part presentation model comprises a plurality of allele-specific models that each receives, as input, the peptide sequences of the plurality of infectious disease-derived antigens, or a representation thereof; and
select a subset of the plurality of infectious disease-derived antigens based on the set of numerical likelihoods to generate a set of selected antigens.
35 - 57 . (canceled)