Machine learning techniques in protein design for vaccine generation

Abstract

One or more data objects defining a plurality of wild-type amino acid sequences are received. A plurality of reduced dimensional sequences are generated in the reduced dimensional space from the one or more data objects. A plurality of candidate sequences are generated in the reduced dimensional space using the plurality of reduced dimensional sequences. One or more data objects defining viral amino acid sequences are received. The viral sequences in the reduced dimensional space are received. Each of the candidate sequences and at least one of the reduced dimensional viral sequences are provided as input to a potency predictor. A candidate score for each of the candidate sequences is received as output from the potency predictor. At least one candidate sequence is selected from among the candidate sequences. At least one new amino acid sequence is generated. Each of the generated amino acid sequences is suitable for producing a respective vaccine.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Application No. 63 / 319,700, filed March 14, 2022, and U.S. Provisional Application No. 63 / 319,692, filed March 14, 2022, the entire contents of which are incorporated by reference herein.

[0002] This application relates to the use of machine learning techniques in vaccine design. [Background technology]

[0003] Machine learning (ML) is the use of computer algorithms that can improve automatically through experience and by using data. Machine learning is considered a part of artificial intelligence. Machine learning algorithms build models based on example data, known as training data, to make predictions or decisions without being explicitly programmed. Machine learning algorithms are used in a wide range of applications where it is difficult or impossible to develop traditional algorithms to perform the required tasks, such as medicine, email filtering, speech recognition, and computer vision.

[0004] A vaccine is a biological product that provides acquired immunity against a specific infectious disease. Vaccines usually contain an agent that resembles the disease-causing microorganism, and are often made from a weakened or killed form of the microorganism, its toxins, or one of its surface proteins. The agent stimulates the body's immune system to recognize the agent as a threat, destroy it, and further recognize and destroy any microorganisms related to the agent that may be encountered in the future. Vaccines can be prophylactic (to prevent or ameliorate the effects of future infections with natural or "wild" pathogens) or therapeutic (to combat diseases that have already developed, such as cancer). Some vaccines provide complete sterilizing immunity, in which infection is completely prevented. Summary of the Invention [Means for solving the problem]

[0005] The strains used in seasonal influenza vaccines are now almost universally selected by public health authorities. These selections are made annually based on observations of immune responses in animal models and human trials. However, H3N2 vaccines using strains recommended by public health authorities have not been sufficient to induce broad protection in the general population, for example over the past five years. Furthermore, published data from this period show that immunological associations are separated into diverse clades, where each clade provides protection for itself, while protection against other clades may be limited. The present disclosure provides a solution to this problem. Implementations described in this disclosure provide an algorithm that can generate influenza (or other) antigens for use as vaccines. In one implementation, this may include: 1) Generate a reduced dimensional space of all wild-type hemagglutinin sequences through machine learning (e.g., a variational autoencoder architecture) that uses the following two steps: a. Variable embedding in a reduced space, e.g., the model predicts the mean and variance from the input sequence and the embedded coordinates are selected from a normal distribution with the predicted mean and variance; b. Then, decoding is performed from the reduced spatial position "autoencoder" loss function back to the original sequence by reducing by the similarity of the input and output sequences. 2) Train an immune response prediction model based on the position of antigens (vaccine candidates) and readout strains (target sequences) in the reduced dimensional space [input: antigens and readouts embedded by the model from step 1, output: measures of immune response such as antibody titers]. 3) Sample candidate vaccine component representations from the reduced space and rank the candidate vaccine component representations by their predictive performance against the target sequence using the model described in step 2 to identify the top candidates. 4) Decode the top candidate representations [using the model from step 1b] to release hemagglutinin sequences that may or may not have been observed in the original wild-type set.

[0006] In one example, an algorithm was used to optimize the HA1 sequence of H3 hemagglutinin (positions 16 to 345), and then the wild-type signal peptide and HA2 region were grafted to create the complete hemagglutinin sequence. An exemplary modified antigen sequence starting with A / Singapore / INFIMH-16-0019 / 2016 is provided with the mutated residues shown in bold: [ka]

[0007] One or more computer systems may be configured to perform certain operations or actions by having software, firmware, hardware, or a combination thereof installed on the system that causes the system to perform the operations during operation. One or more computer programs may be configured to perform certain operations or actions by including instructions that, when executed by a data processing device, cause the device to perform the operations. One general aspect includes a dimensionality reduction method for generating amino acid sequences, the method being performed by one or more computer systems. The method includes receiving one or more data objects defining a plurality of wild-type amino acid sequences. The method also includes generating a plurality of reduced dimensional arrays in a reduced dimensional space from the one or more data objects, each reduced dimensional array including data for at least one respective wild-type amino acid sequence, the reduced dimensional space being lower dimensional than the wild-type amino acid sequence, and the plurality of reduced dimensional arrays defining a distribution of values ​​along each dimension of the reduced dimensional space. The method also includes generating a plurality of candidate sequences in the reduced dimensional space using the plurality of reduced dimensional arrays. The method also includes receiving one or more data objects defining a viral amino acid sequence. The method also includes generating at least one reduced dimensional viral sequence in the reduced dimensional space. The method also includes providing each of the candidate sequences and the at least one reduced dimension viral sequence as input to a potency predictor. The method also includes receiving a candidate score for each of the candidate sequences as output from the potency predictor. The method also includes selecting at least one candidate sequence from among the candidate sequences. The method also includes generating at least one new amino acid sequence for each of the selected candidate sequences. The method also includes providing the generated at least one amino acid sequence.The method also includes operations where each of the generated amino acid sequences is suitable for producing a respective vaccine, which may include at least one of the following groups: i) a protein defined by the generated amino acid sequence, ii) a nucleic acid capable of producing the protein defined by the generated amino acid sequence, and iii) a delivery vehicle capable of producing the protein defined by the generated amino acid sequence. Other embodiments of this aspect include corresponding computer systems, devices, and computer programs stored on one or more computer storage devices, each configured to perform the operations of the method.

[0008] Implementations may include one or more of the following features. The method includes operations in which generating a plurality of reduced dimensional arrays may include creating a representation of a wild-type amino acid sequence using a variational autoencoder that predicts mean and variance values ​​of input data. Each of the reduced dimensional arrays may include a respective set of values, and generating a plurality of candidate sequences in the reduced dimensional space may include sampling a distribution of values ​​of the plurality of reduced dimensional arrays. The potency predictor is configured to receive as input: i) a first sequence in the reduced dimensional space, and ii) a second sequence in the reduced dimensional space; and to provide as output a potency score as a candidate score, the potency score defining a measure of a biological response between the first sequence and the second sequence. Selecting at least one candidate sequence as a selected candidate sequence may include selecting n candidate sequences having the highest candidate scores. The method includes operations in which n is a value of 1 such that a single candidate sequence is selected. The method includes operations in which n is a value greater than 1 such that a plurality of candidate sequences are selected. Selecting at least one candidate sequence as a selected candidate sequence may include selecting candidate sequences whose respective candidate scores are greater than a threshold value. Each of the generated amino acid sequences is different from any of the wild type amino acid sequences. At least one of the candidate sequences is within the plurality of reduced dimension sequences. Each vaccine is for one of a group that may include i) influenza, ii) human rhinovirus, iii) HIV, and iiiv) coronavirus diseases. Implementations of the described technology may include hardware, methods or processes, or computer software on a computer accessible medium.

[0009] One general aspect includes a system for generating amino acid sequences, which may include a computer memory. The system also includes one or more processors. The system also includes a computer memory that stores instructions that, when executed by the processor, cause the processor to perform operations, including receiving one or more data objects defining a plurality of wild-type amino acid sequences; generating a plurality of reduced-dimensional sequences in a reduced-dimensional space from the one or more data objects, where each reduced-dimensional sequence contains data for at least one respective wild-type amino acid sequence, the reduced-dimensional space being lower dimensional than the wild-type amino acid sequence, and the plurality of reduced-dimensional sequences defines a distribution of values ​​along each dimension of the reduced-dimensional space; generating a plurality of candidate sequences in the reduced-dimensional space using the plurality of reduced-dimensional sequences; receiving one or more data objects defining a viral amino acid sequence; generating at least one reduced-dimensional viral sequence in the reduced-dimensional space. providing each of the candidate sequences and at least one of the reduced dimensional viral sequence as input to a titer predictor; receiving a candidate score for each of the candidate sequences as output from the titer predictor; selecting at least one candidate sequence from among the candidate sequences; generating at least one new amino acid sequence for each of the selected candidate sequences; and providing the generated at least one amino acid sequence, wherein each of the generated amino acid sequences is suitable for producing a respective vaccine comprising at least one of the group consisting of i) a protein defined by the generated amino acid sequence, ii) a nucleic acid capable of producing the protein defined by the generated amino acid sequence, and iii) a delivery vehicle capable of producing the protein defined by the generated amino acid sequence. Other embodiments of this aspect include corresponding computer systems, devices, and computer programs stored on one or more computer storage devices, each configured to perform the operations of the method.

[0010] Implementations may include one or more of the following features: The system, where the generation of the multiple reduced dimensional arrays may include creating a representation of the wild-type amino acid sequence using a variational autoencoder that predicts the mean and variance values ​​of the input data. Each of the reduced dimensional arrays may include a respective group of values, and the generation of the multiple candidate sequences in the reduced dimensional space may include sampling a distribution of values ​​of the multiple reduced dimensional arrays. The potency predictor is configured to receive as input: i) a first sequence in the reduced dimensional space, and ii) a second sequence in the reduced dimensional space; and to provide as output a potency score as the candidate score, the potency score defining a measure of a biological response between the first sequence and the second sequence. Selecting at least one candidate sequence as the selected candidate sequence may include selecting the n candidate sequences with the highest candidate scores. Implementations of the described technology may include hardware, methods or processes, or computer software on a computer-accessible medium.

[0011] One general aspect includes a non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform operations including: receiving one or more data objects defining a plurality of wild-type amino acid sequences; generating a plurality of reduced-dimensional arrays in a reduced-dimensional space from the one or more data objects, where each reduced-dimensional array contains data for at least one respective wild-type amino acid sequence, the reduced-dimensional space being lower dimensional than the wild-type amino acid sequence, and the plurality of reduced-dimensional arrays defining a distribution of values ​​along each dimension of the reduced-dimensional space; generating a plurality of candidate sequences in the reduced-dimensional space using the plurality of reduced-dimensional arrays; receiving one or more data objects defining viral amino acid sequences; generating at least one reduced-dimensional viral sequence in the reduced-dimensional space. The method may include generating a sequence of the candidate sequences; providing each of the candidate sequences and at least one of the reduced dimensional viral sequence as input to a potency predictor; receiving a candidate score for each of the candidate sequences as output from the potency predictor; selecting at least one candidate sequence from among the candidate sequences; generating at least one new amino acid sequence for each of the selected candidate sequences; and providing the generated at least one amino acid sequence, wherein each of the generated amino acid sequences is suitable for producing a respective vaccine comprising at least one of the group consisting of i) a protein defined by the generated amino acid sequence, ii) a nucleic acid capable of producing the protein defined by the generated amino acid sequence, and iii) a delivery vehicle capable of producing the protein defined by the generated amino acid sequence. Other embodiments of this aspect include corresponding computer systems, devices, and computer programs stored on one or more computer storage devices, each configured to perform the operations of the method.

[0012] Implementations may include one or more of the following features: A medium in which generating a plurality of reduced dimensional arrays may include creating a representation of a wild-type amino acid sequence using a variational autoencoder that predicts mean and variance values ​​of input data. Each of the reduced dimensional arrays may include a set of respective values, and generating a plurality of candidate sequences in the reduced dimensional space may include sampling a distribution of values ​​of the plurality of reduced dimensional arrays. The potency predictor is configured to receive as input: i) a first sequence in the reduced dimensional space, and ii) a second sequence in the reduced dimensional space; and to provide as output a potency score as a candidate score, the potency score defining a measure of a biological response between the first sequence and the second sequence. Implementations of the described technology may include hardware, a method or process, or computer software on a computer-accessible medium.

[0013] Also disclosed herein are vaccine compositions comprising a plurality of any of the amino acid sequences generated by the methods described herein.

[0014] Also disclosed are vectors, fusion proteins, and cells that include one or more of the peptides and / or proteins produced according to the methods described herein.

[0015] Also disclosed herein are methods of eliciting an immune response in a subject comprising administering one or more of the isolated nucleic acids, peptides, and / or proteins described herein, thereby eliciting an immune response in a subject.

[0016] In one aspect, disclosed herein is a method of inhibiting viral infection comprising administering to a subject one or more of the isolated nucleic acids, peptides and / or proteins described herein, or a vaccine comprising any of the isolated nucleic acids, peptides and / or proteins described herein.

[0017] Also disclosed herein is a method of immunizing a subject against influenza virus, comprising administering to the subject an immunologically effective amount of a vaccine composition disclosed herein. Also disclosed herein is a vaccine composition disclosed herein for use in a method of immunizing a subject against a virus (e.g., influenza virus). Also disclosed herein is a vaccine composition disclosed herein for manufacture of a medicament for use in a method of immunizing a subject against a virus (e.g., influenza virus). In certain embodiments, the method prevents a viral infection (e.g., influenza virus infection) in the subject, and in certain embodiments, the method generates a protective immune response (e.g., an HA antibody response and / or an NA antibody response) in the subject. In certain embodiments, the subject is a human, and in certain embodiments, the vaccine composition is administered intramuscularly, intradermally, subcutaneously, intravenously, or intraperitoneally.

[0018] Another aspect of the present disclosure relates to a method for reducing one or more symptoms of a viral infection (e.g., influenza virus infection), comprising administering to a subject a prophylactically effective amount of a vaccine composition disclosed herein. Also disclosed herein is a vaccine composition disclosed herein for use in a method for reducing one or more symptoms of a viral infection (influenza virus infection). Also disclosed herein is a vaccine composition disclosed herein for the manufacture of a medicament for use in a method for reducing one or more symptoms of a viral infection (influenza virus infection).

[0019] In various embodiments, the methods and compositions disclosed herein treat or prevent disease caused by either or both seasonal or pandemic virus strains (e.g., either or both seasonal and pandemic influenza strains).

[0020] In embodiments of the methods disclosed herein where the subject is a human, the human is at least 6 months old, under 18 years old, at least 6 months old and under 18 years old, at least 18 years old and under 65 years old, at least 6 months old and under 5 years old, at least 5 years old and under 65 years old, at least 60 years old, or at least 65 years old. For example, the subject is 6 months old, 8 months old, 10 months old, 12 months old, 14 months old, 16 months old, 18 months old, 20 months old, 22 months old, 24 months old, 3 years old, 4 years old, 5 years old, 6 years old, 10 years old, 12 years old, 15 years old, 18 years old, 20 years old, 21 years old, 25 years old, 30 years old, 35 years old, 40 years old, 50 years old, 60 years old, 70 years old, 75 years old, 80 years old, 85 years old, or 90 years old. In certain embodiments, the methods disclosed herein comprise administering to a subject two doses of a vaccine composition, spaced 2-6 weeks apart, for example, 4 weeks apart.

[0021] The implementations discussed in this disclosure may provide one or more of the following advantages: The implementations may be used to generate hemagglutinin sequences that may induce broad protection from influenza infection after vaccination. In particular, the implementations may be used to produce antigens with higher than expected recovery rates of functional influenza viruses with the designed hemagglutinin sequences. These antigens are believed to have broader protection in animal models than current standard of care antigens. The implementations may be used to generate broadly protective hemagglutinin proteins for use as influenza vaccine antigens.

[0022] By converting amino acid sequences into a low-dimensional space and making changes to their representation, novel non-wild type amino acid sequences can be generated. These new amino acid sequences can be used to produce new, more effective vaccines than would otherwise be possible. For example, there may be non-wild type amino acid sequences that generate a stronger immune response in a subject (e.g., a human) when exposed to a vaccine having a protein with the non-wild type amino acid sequence, resulting in greater protection and greater health.

[0023] Another advantage of the techniques provided in this disclosure is that they improve the likelihood of generating protein sequence data for proteins that may actually exist and be manufactured. As will be appreciated, it is possible to describe protein sequences that are not possible due to geometry, physical forces, etc. The processes described in this document can be advantageously limited to only those processes that are known or expected to be manufacturable.

[0024] Other features, aspects, and potential advantages will be apparent from the accompanying description and figures. [Brief description of the drawings]

[0025] [Figure 1] FIG. 1 is a block diagram of an example system that can be used to produce a vaccine. [Diagram 2] FIG. 1 is a schematic representation of data that can be used in the manufacture of a vaccine. [Diagram 3] 1 is a flowchart of an example process that can be used to process high dimensional data in a low dimensional space, such as may be used in the production of a vaccine. [Figure 4] 1 is a flowchart of an example process that can be used to process high dimensional data in a low dimensional space, such as may be used in the production of a vaccine. [Diagram 5] 1 is a flowchart of an example process that can be used to process high dimensional data in a low dimensional space, such as may be used in the production of a vaccine. [Figure 6] 1 is a flowchart of an example process that can be used to process high dimensional data in a low dimensional space, such as may be used in the production of a vaccine. [Figure 7] FIG. 1 is a swim lane diagram of an example process for producing a vaccine. [Figure 8] FIG. 1 is a schematic diagram illustrating an example of a computing device and a mobile computing device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] Reference characters in the various drawings indicate like elements.

[0027] Vaccine creation by machine learning processes is described herein. Vaccine creation uses candidate proteins generated by computational processes including machine learning. A corpus of wild-type amino acid sequences is provided to a variable autoencoder to generate low-dimensional representations (latent space) of sequences. After training of such a model, some representations may generate non-wild-type amino acid sequences when decoded. The representations in the latent space are tested to identify one or more representations that are computationally predicted to generate amino acid sequences that generate a desired biological response in a subject. One or more of these candidate representations are selected based on the predicted desired response and transformed into the high-dimensional space of traditional amino acid sequence definitions. Vaccines are produced from these newly defined amino acid sequences. If necessary, the sequences can be filtered to exclude non-wild-type or wild-type sequences.

[0028] Influenza viruses are members of the Orthomyxoviridae family. There are three subtypes of influenza viruses: influenza A, influenza B, and influenza C. Influenza A viruses infect a wide variety of birds and mammals, including humans, chickens, ferrets, pigs, and horses. In mammals, most influenza A viruses cause mild localized infections of the respiratory and intestinal tracts.

[0029] Influenza virions contain a negative-sense RNA genome that encodes nine proteins: hemagglutinin (HA), matrix (M1), proton ion channel protein (M2), neuraminidase (NA), nonstructural protein 2 (NS2), nucleoprotein (NP), polymerase acidic protein (PA), polymerase basic protein 1 (PB1), and polymerase basic protein 2 (PB2). HA, M1, M2, and NA are membrane-associated proteins, while NP, NS2, PA, PB1, and PB2 are nucleocapsid-associated proteins. M1 protein is the most abundant protein in influenza particles. HA and NA proteins are envelope glycoproteins involved in virus attachment and cell entry. HA and NA proteins are the source of the major immunodominant epitopes for virus neutralization and protective immunity. HA and NA proteins are considered the most important components of prophylactic influenza vaccines.

[0030] HA is a viral surface glycoprotein that generally contains approximately 560 amino acids and accounts for 25% of the total viral protein.

[0031] NA is a membrane glycoprotein of influenza viruses. NA is 413 amino acids in length and is encoded by a gene of 1413 nucleotides. Nine different NA subtypes have been identified in influenza viruses (N1, N2, N3, N4, N5, N6, N7, N8, and N9), all of which are found in wild birds.

[0032] The ability of influenza viruses to cause a wide range of diseases derives from their ability to evade the immune system by undergoing antigenic change.

[0033] definition In order that this disclosure may be more readily understood, certain terms are first defined below. Throughout the specification, additional definitions of the following terms and other terms may be set forth. In the event that a definition of a term set forth below conflicts with a definition in an application or patent incorporated by reference, the meaning of the term shall be understood using the definition set forth in this application.

[0034] As used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. So, for example, reference to "a method" includes one or more methods and / or steps of the type described herein and / or that will become apparent to those skilled in the art after reading this disclosure and so forth.

[0035] Adjuvant: As used herein, the term "adjuvant" refers to a substance or combination of substances that can be used to enhance the immune response to the antigenic component of a vaccine.

[0036] Antigen: As used herein, the term "antigen" refers to an agent that elicits an immune response when exposed to or administered to an organism, and / or (ii) an agent that binds to a T cell receptor (e.g., when presented by an MHC molecule) or an antibody (e.g., produced by a B cell). In some embodiments, the antigen elicits a humoral response in the organism (e.g., including production of antigen-specific antibodies); alternatively or additionally, in some embodiments, the antigen elicits a cellular response in the organism (e.g., involving T cells whose receptors specifically interact with the antigen). One of skill in the art will appreciate that a particular antigen may elicit an immune response in one or more members of a target organism (e.g., mice, ferrets, rabbits, primates, humans), but not in all members of the target organism's species. In some embodiments, the antigen elicits an immune response in at least about 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% of members of the target species. In some embodiments, the antigen binds to an antibody and / or a T cell receptor and may or may not elicit a specific physiological response in the organism. In some embodiments, for example, the antigen may bind to an antibody and / or a T cell receptor in vitro, regardless of whether such interactions occur in vivo. In some embodiments, the antigen reacts with the products of specific humoral or cellular immunity, including those induced by heterologous immunogens. Antigens include forms of NA and HA as described herein.

[0037] Carrier: As used herein, the term "carrier" refers to a diluent, adjuvant, excipient, or vehicle with which the composition is administered. In some exemplary embodiments, the carrier may include sterile liquids, such as water and oils, including those of petroleum, animal, vegetable, or synthetic origin, such as peanut oil, soybean oil, mineral oil, sesame oil, and the like. In some embodiments, the carrier is or includes one or more solid ingredients.

[0038] Epitope: As used herein, the term "epitope" includes any moiety that is specifically recognized, in whole or in part, by an immunoglobulin (e.g., antibody or receptor) binding component. In some embodiments, an epitope is composed of multiple chemical atoms or groups on an antigen. In some embodiments, such chemical atoms or groups are surface exposed when the antigen adopts a relevant three-dimensional structure. In some embodiments, such chemical atoms or groups are physically close to each other in space when the antigen adopts such a structure. In some embodiments, at least some of such chemical atoms or groups are physically separated from each other when the antigen adopts an alternative structure (e.g., linearized).

[0039] Excipient: As used herein, the term "excipient" refers to a non-therapeutic agent that may be included in a pharmaceutical composition, for example, to impart or contribute to a desired consistency or stabilizing effect. Suitable pharmaceutical excipients include, for example, starch, glucose, lactose, sucrose, sorbitol, gelatin, malt, rice, flour, chalk, silica gel, sodium stearate, glycerol monostearate, talc, sodium chloride, dried skim milk, glycerol, propylene, glycol, water, ethanol, and the like.

[0040] Immune response: As used herein, the term "immune response" refers to the reaction of cells of the immune system, such as B cells, T cells, dendritic cells, macrophages, or polymorphonuclear cells, to a stimulus, such as an antigen, immunogen, or vaccine. An immune response can include any cell of the body involved in a host defense response, including, for example, epithelial cells that secrete interferons or cytokines. Immune responses include, but are not limited to, innate and / or adaptive immune responses. As used herein, a protective immune response refers to an immune response that protects a subject from infection (prevents infection or prevents the development of a disease associated with an infection) or reduces the symptoms of an infection. Methods for measuring immune responses are well known in the art and include, for example, measuring the proliferation and / or activity of lymphocytes (such as B cells or T cells), measuring the secretion of cytokines or chemokines, measuring inflammation, measuring antibody production, and the like. An antibody response or humoral response is an immune response in which antibodies are produced. A "cellular immune response" is one that is mediated by T cells and / or other white blood cells.

[0041] Immunogen: As used herein, the term "immunogen" or "immunogenic" refers to a compound, composition, or substance that, under appropriate conditions, is capable of stimulating an immune response, such as the production of antibodies or a T cell response, in an animal, including compositions that are injected or absorbed into an animal. As used herein, "immunize" means to achieve a subject protected from infectious disease.

[0042] Immunologically effective amount: As used herein, the term "immunologically effective amount" means an amount sufficient to immunize a subject.

[0043] Prevention: The term "prevention," as used herein, refers to preventing, avoiding disease onset, delaying the onset, and / or reducing the frequency and / or severity of one or more symptoms of a particular disease, disorder, or condition (e.g., infection with an influenza virus). In some embodiments, prevention is assessed on a population basis, and an agent is considered to "prevent" a particular disease, disorder, or condition if a statistically significant reduction in the occurrence, frequency, and / or intensity of one or more symptoms of the disease, disorder, or condition is observed in a population susceptible to the disease, disorder, or condition.

[0044] Sequence identity: The similarity between amino acid or nucleic acid sequences is expressed in terms of the similarity between the sequences, otherwise referred to as sequence identity. Sequence identity is often measured in terms of the percentage of identity (or similarity or homology); the higher the percentage, the more similar the two sequences are. "Sequence identity" between two nucleic acid sequences indicates the percentage of nucleotides that are identical between the sequences. "Sequence identity" between two amino acid sequences indicates the percentage of amino acids that are identical between the sequences. Homologs or variants of a given gene or protein have a relatively high degree of sequence identity when aligned using standard methods.

[0045] The terms "% identical", "% identity" or similar terms are intended to refer in particular to the percentage of nucleotides or amino acids that are identical in optimal alignment between the sequences to be compared. Said percentage is purely statistical, and the differences between the two sequences may, but do not necessarily, be randomly distributed over the entire length of the sequences to be compared. Comparison of two sequences is usually performed by comparing said sequences over segments or "windows of comparison" after optimal alignment in order to identify local regions of corresponding sequences. Optimal alignment for comparison can be performed manually or using the local homology algorithm of Smith and Waterman, 1981, Ads App. Math. 2, 482, using the local homology algorithm of Needleman and Wunsch, 1970, J. Mol. Biol. 48, 443, using the similarity search algorithm of Pearson and Lipman, 1988, Proc. Natl Acad. Sci. USA 88, 2444, or using computer programs that use the above algorithms (GAP, BESTFIT, FASTA, BLAST P, BLAST N and TFASTA from the Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Drive, Madison, Wis.).

[0046] The percentage of identity is obtained by determining the number of corresponding identical positions in the sequences to be compared, dividing this number by the number of positions being compared (e.g., the number of positions in the reference sequence) and multiplying this result by 100.

[0047] In some embodiments, the degree of identity is given for a region that is at least about 50%, at least about 60%, at least about 70%, at least about 80%, at least about 90%, or about 100% of the entire length of the reference sequence. For example, if the reference nucleic acid sequence consists of 200 nucleotides, the degree of identity is given for at least about 100, at least about 120, at least about 140, at least about 160, at least about 180, or about 200 nucleotides, in some embodiments, in consecutive nucleotides. In some embodiments, the degree of identity is given for the entire length of the reference sequence.

[0048] A nucleic acid sequence or amino acid sequence that has a particular degree of identity to a given nucleic acid sequence or amino acid sequence, respectively, may have at least one functional and / or structural characteristic of the given sequence, e.g., in some instances, is functionally and / or structurally equivalent to the given sequence. In some embodiments, a nucleic acid sequence or amino acid sequence that has a particular degree of identity to a given nucleic acid sequence or amino acid sequence is functionally and / or structurally equivalent to the given sequence.

[0049] Subject: As used herein, the term "subject" refers to any member of the animal kingdom. In some embodiments, "subject" refers to a human. In some embodiments, "subject" refers to a non-human animal. In some embodiments, subjects include, but are not limited to, mammals, birds, reptiles, amphibians, fish, insects, and / or worms. In some embodiments, the non-human subject is a mammal (e.g., a rodent, mouse, rat, rabbit, ferret, monkey, dog, cat, sheep, cow, primate, and / or pig). In some embodiments, the subject may be a transgenic animal, a genetically engineered animal, and / or a clone. In some embodiments, the subject is an adult, an adolescent, or an infant. In some embodiments, the terms "individual" or "patient" are used and are intended to be interchangeable with "subject."

[0050] Vaccination: As used herein, the term "vaccination" or "vaccinate" refers to the administration of a composition intended to generate an immune response against a disease-causing agent, such as, for example, an influenza virus. Vaccination can be administered before, during, and / or after exposure to the disease-causing agent, and / or before, during, and / or after the onset of one or more symptoms, in some embodiments before, during, and / or immediately after exposure to the agent. Vaccines can induce both prophylactic (preventative) and therapeutic responses. Methods of administration vary depending on the vaccine, but may include inoculation, ingestion, inhalation, or other forms of administration. Inoculation may be delivered by any of a number of routes, including parenterally, such as intravenously, subcutaneously, intraperitoneally, intradermally, or intramuscularly. Vaccines may be administered with adjuvants to enhance the immune response. In some embodiments, vaccination includes multiple administrations of the vaccination composition at appropriate time intervals.

[0051] Vaccine efficacy: As used herein, the term "vaccine efficacy" or "vaccine efficacy" refers to a measure of the percentage reduction in evidence of disease among subjects administered a vaccine composition. For example, 50% vaccine efficacy indicates a 50% reduction in the number of disease cases in a group of vaccinated subjects compared to a group of unvaccinated subjects or a group of subjects administered a different vaccine.

[0052] Wild-type (WT): As understood in the art, the term "wild-type" generally refers to the normal form of a protein or nucleic acid as found in nature. For example, wild-type HA and NA polypeptides are found in natural isolates of influenza viruses. A variety of different wild-type HA and NA sequences can be found in the NCBI influenza virus sequence database.

[0053] Measurement of hemagglutinin activity Hemagglutinin activity may be measured using techniques known in the art, such as the hemagglutinin inhibition assay (HAI). HAI applies a process of hemagglutination, called hemagglutination, in which sialic acid receptors on the surface of red blood cells (RBCs) bind to the hemagglutinin glycoprotein found on the surface of influenza viruses (and some other viruses), creating a network or lattice structure of interconnected RBCs and virus particles that occurs in a concentration-dependent manner on the virus particles. This is a physical measurement taken as a proxy for the ability of the virus to bind to similar sialic acid receptors on cells that target pathogens in the body. The introduction of anti-viral antibodies, generated in a human or animal immune response to another virus (which may be genetically similar or different from the virus used to bind to the RBCs in the assay), alters the concentration of the virus enough to disrupt the virus-RBC interaction and change the concentration at which hemagglutination is observed in the assay. One goal of HAI may be to characterize the concentration of antibodies in an antiserum or other sample that contains the antibodies, relative to their ability to inhibit hemagglutination in the assay. The highest dilution of antibody that prevents hemagglutination is called the HAI titer (ie, the measured response).

[0054] Another approach to measure HA antibody responses is to measure a potentially larger set of antibodies elicited by human or animal immune responses, which may not necessarily affect hemagglutination in HAI assays. A common approach for this is to utilize enzyme-linked immunosorbent assay (ELISA) technology, in which a viral antigen (e.g., hemagglutinin) is immobilized on a solid surface and then antibodies from antisera are allowed to bind to the antigen. The readout measures the catalysis of a substrate for an exogenous enzyme conjugated to an antibody from the antisera, or to another antibody that itself binds to the antibody of the antisera. The catalysis of the substrate produces a readily detectable product. There are many variations of this type of in vitro assay. One such variation is called antibody forensics (AF), which is a multiplex bead array technology that allows a single serum sample to be measured simultaneously against many antigens. These measurements characterize concentration and total antibody recognition when compared to HAI titers, which are believed to be more specifically related to interference with sialic acid binding by the hemagglutinin molecule. Thus, antisera antibodies may, in some cases, have a proportionally higher or lower measurement than the corresponding HAI titer for the hemagglutinin molecule of one virus compared to the hemagglutinin molecule of another virus; in other words, these two measurements, AF and HAI, are not linearly correlated.

[0055] Another method for measuring HA antibody responses includes virus neutralization assays (e.g., microneutralization assays), in which antibody titers are measured in permissive cell cultures after incubation of virus with serial dilutions of antibody / serum samples by the reduction in plaques, foci, and / or fluorescent signal, depending on the specific neutralization assay technique.

[0056] Measurement of neuraminidase activity Neuraminidase activity may be measured using techniques known in the art, including, for example, the MUNANA assay, the ELLA assay, or the NA-Star® assay (ThermoFisher Scientific, Waltham, Mass.). The MUNANA assay uses 2'-(4-methylumbelliferyl)-alpha-DN-acetylneuraminic acid (MUNANA) as a substrate. Any enzymatically active neuraminidase in the sample will cleave the MUNANA substrate, releasing the fluorescent compound 4-methylumbelliferone (4-MU). Thus, the amount of neuraminidase activity in the test sample correlates with the amount of 4-MU released, which may be measured using fluorescence intensity (RFU, relative fluorescence units).

[0057] For the purpose of determining the neuraminidase activity of the soluble tetrameric NA of the present disclosure, the MUNANA assay should be performed using the following conditions: mix the soluble tetrameric NA with buffer [33.3 mM 2-(N-morpholino)ethanesulfonic acid (MES, pH 6.5), 4 mM CaCl2, 50 mM BSA] and substrate (100 μM MUNANA) and incubate at 37 ° C for 1 hour with shaking; stop the reaction by adding alkaline pH solution (0.2 M Na2CO3); measure the fluorescence intensity using an excitation wavelength of 355 nm and an emission wavelength of 460 nm, respectively; calculate the enzyme activity against a 4MU reference. If necessary, an equivalent assay can be used to measure neuraminidase enzyme activity.

[0058] Vaccine composition: In certain aspects, disclosed herein are vaccine compositions comprising a plurality of engineered amino acid sequences.

[0059] Each generated amino acid sequence may be present in the composition disclosed herein in an amount effective to induce an immune response in a subject to which the composition is administered. In certain embodiments, each generated amino acid sequence may be present in the vaccine composition disclosed herein in an amount ranging from about 0.1 g to about 500 g, such as about 5 g to about 120 g, about 1 g to about 60 g, about 10 g to about 60 g, about 15 g to about 60 g, about 40 g to about 50 g, about 42 g to about 47 g, about 5 g to about 45 g, about 15 g to about 45 g, about 0.1 g to about 90 g, about 5 g to about 90 g, about 10 g to about 90 g, or about 15 g to about 90 g. In certain embodiments, each recombinant HA may be present in a vaccine composition disclosed herein in an amount of about 5 g, 10 g, 15 g, 20 g, 25 g, 30 g, 35 g, 40 g, 45 g, 50 g, 55 g, 60 g, 65 g, 70 g, 75 g, 80 g, 85 g, or about 90 g.

[0060] The vaccine composition may also further comprise an adjuvant. As used herein, the term "adjuvant" refers to a substance or vehicle that non-specifically enhances the immune response to an antigen. Adjuvants may include inorganic substances (alum, aluminum salts (e.g., aluminum hydroxide / oxyhydroxide (AlOOH), aluminum phosphate (AlPO4), aluminum hydroxyphosphate sulfate (AAHS), and / or aluminum potassium sulfate)) suspensions to which antigens are adsorbed; or water-in-oil emulsions in which antigen solutions are emulsified in mineral oil (e.g., Freund's incomplete adjuvant), which may also include killed mycobacteria (Freund's complete adjuvant) to further enhance antigenicity. Immunostimulatory oligonucleotides (e.g., those that contain CpG motifs) can also be used as adjuvants (see, for example, U.S. Patent Nos. 6,194,388; 6,207,646; 6,214,806; 6,218,371; 6,239,116; 6,339,068; 6,406,705; and 6,429,199).Adjuvants also include biological molecules such as lipids and costimulatory molecules. Exemplary biological adjuvants include AS04 (Didierlaurent, AM et al, AS04, an Aluminum Salt- and TLR4 Agonist-Based Adjuvant System, Induces a Transient Localized Innate Immune Response Leading to Enhanced Adaptive Immunity, J. IMMUNOL. 2009, 183:6186-6197), IL-2, RANTES, GM-CSF, TNF-?, IFN-?, G-CSF, LFA-3, CD72, B7-1, B7-2, OX-40L, and 41 BBL.

[0061] In certain embodiments, the adjuvant is a squalene-based adjuvant comprising an oil-in-water adjuvant emulsion comprising at least squalene, an aqueous solvent, a polyoxyethylene alkyl ether hydrophilic non-ionic surfactant, and a hydrophobic non-ionic surfactant. In certain embodiments, the emulsion is thermoreversible, and optionally has 90% of the population by volume of oil droplets having a size of less than 200 nm.

[0062] In certain embodiments, the polyoxyethylene alkyl ether is of the formula CH3-(CH2)x-(O-CH2-CH2)n-OH, where n is an integer from 10 to 60 and x is an integer from 11 to 17. In certain embodiments, the polyoxyethylene alkyl ether surfactant is polyoxyethylene (12) cetostearyl ether.

[0063] In certain embodiments, 90% of the population by volume of the oil droplets have a size of less than 160 nm. In certain embodiments, 90% of the population by volume of the oil droplets have a size of less than 150 nm. In certain embodiments, 50% of the population by volume of the oil droplets have a size of less than 100 nm. In certain embodiments, 50% of the population by volume of the oil droplets have a size of less than 90 nm.

[0064] In certain embodiments, the adjuvant further comprises at least one alditol, including, but not limited to, glycerol, erythritol, xylitol, sorbitol, and mannitol.

[0065] In certain embodiments, the hydrophilic nonionic surfactant has a hydrophilic lipophilic balance (HLB) of 10 or more. In certain embodiments, the hydrophobic nonionic surfactant has an HLB of less than 9. In certain embodiments, the hydrophilic nonionic surfactant has an HLB of 10 or more and the hydrophobic nonionic surfactant has an HLB of less than 9.

[0066] In certain embodiments, the hydrophobic non-ionic surfactant is a sorbitan ester, such as sorbitan monooleate, or a mannide ester surfactant. In certain embodiments, the amount of squalene is 5-45%. In certain embodiments, the amount of polyoxyethylene alkyl ether surfactant is 0.9-9%. In certain embodiments, the amount of hydrophobic non-ionic surfactant is 0.7-7%. In certain embodiments, the adjuvant comprises i) 32.5% squalene, ii) 6.18% polyoxyethylene (12) cetostearyl ether, iii) 4.82% sorbitan monooleate, and iv) 6% mannitol.

[0067] In certain embodiments, the adjuvant further comprises an alkyl polyglycoside and / or a cryoprotectant, such as a sugar, in particular dodecyl maltoside and / or sucrose.

[0068] In certain embodiments, the adjuvant comprises AF03 as described in Klucker et al., AF03, an alternative squalene emulsion-based vaccine adjuvant prepared by a phase inversion temperature method, J. PHARM. SCI. 2012, 101(12):4490-4500, which is incorporated herein by reference in its entirety. In certain embodiments, the adjuvant comprises a liposome-based adjuvant such as SPA14, which is a liposome-based adjuvant (AS01-like) that comprises a Toll-like receptor 4 (TLR4) agonist (E6020) and a saponin (QS21).

[0069] In addition to the recombinant HA, recombinant NA, and optional adjuvants, the vaccine composition may also further comprise one or more pharma- ceutically acceptable excipients. In general, the nature of the excipient will depend on the particular mode of administration used. For example, parenteral formulations usually contain an injectable fluid, which contains pharma- ceutically and physiologically acceptable fluids, such as water, physiological saline, balanced salt solutions, aqueous dextrose, glycerol, and the like, as a vehicle. In the case of solid compositions (e.g., powder, pill, tablet, or capsule forms), conventional non-toxic solid carriers may include, for example, pharmaceutical grades of mannitol, lactose, starch, or magnesium stearate. In addition to biologically neutral carriers, the vaccine composition to be administered may contain minor amounts of non-toxic auxiliary substances, such as wetting or emulsifying agents, pharma- ceutically acceptable salts for adjusting osmotic pressure, preservatives, stabilizers, buffers, sugars, amino acids, pH buffers, and the like (e.g., sodium acetate or sorbitan monolaurate).

[0070] Typically, the vaccine composition is a sterile liquid solution formulated for parenteral administration, such as intravenous, subcutaneous, intraperitoneal, intradermal, or intramuscular. The vaccine composition may also be formulated for intranasal or inhalation administration. The vaccine composition may also be formulated for any other intended route of administration.

[0071] In some embodiments, the vaccine composition is formulated for intradermal injection, intranasal administration, or intramuscular injection. In some embodiments, the injectables are prepared in conventional forms, either as liquid solutions or suspensions, as solid forms suitable for solution or suspension in liquid prior to injection, or as emulsions. In some embodiments, the injectable solutions and suspensions are prepared from sterile powders or granules. General considerations in the formulation and manufacture of pharmaceuticals for administration by these routes can be found, for example, in Remington's Pharmaceutical Sciences, 19th ed., Mack Publishing Co., Easton, PA, 1995, which is incorporated herein by reference. Currently, oral or nasal spray or aerosol routes (e.g., by inhalation) are most commonly used to deliver therapeutics directly to the lungs and respiratory system. In some embodiments, the vaccine composition is administered using a device that delivers a metered amount of the vaccine composition. Suitable devices for use in delivering the intradermal pharmaceutical compositions described herein include short needle devices, such as those described in U.S. Patent Nos. 4,886,499, 5,190,521, 5,328,483, 5,527,288, 4,270,537, 5,015,235, 5,141,496, and 5,417,662 (all of which are incorporated herein by reference).Intradermal compositions can also be administered by devices that limit the effective penetration length of the needle into the skin, such as those described in WO 1999 / 34850 (which are incorporated herein by reference) and their functional equivalents. Also suitable are jet injection devices that deliver liquid vaccines to the dermis by means of a liquid jet injector or by means of a needle that pierces the stratum corneum and produces a jet that reaches the dermis.Jet injection devices are described, for example, in U.S. Pat. Nos. 5,480,381, 5,599,302, 5,334,144, 5,993,412, 5,649,912, 5,569,189, 5,704,911, 5,383,851, 5,893,397, 5,466,220, 5,339,163, and 5,31 The method is described in WO 2006 / 02335, WO 2006 / 02335, WO 2005 ...

[0072] Preparations for parenteral administration typically include sterile aqueous or non-aqueous solutions, suspensions, and emulsions. Examples of non-aqueous solvents include propylene glycol, polyethylene glycol, vegetable oils such as olive oil, and injectable organic esters such as ethyl oleate. Aqueous carriers include water, alcoholic / aqueous solutions, emulsions, or suspensions (e.g., saline and buffered media). Parenteral vehicles include sodium chloride solution, Ringer's dextrose, dextrose, and sodium chloride, lactated Ringer's, or fixed oils. Intravenous vehicles include fluid and nutrient replenishers, electrolyte replenishers (e.g., those based on Ringer's dextrose), and the like. Preservatives and other additives may also be present, such as, for example, antimicrobials, antioxidants, chelating agents, and inert gases, and the like.

[0073] kit Further disclosed herein are kits for the vaccine compositions disclosed herein, which may include a suitable container containing the vaccine composition or multiple containers containing different components of the vaccine composition, optionally together with instructions for use.

[0074] In certain embodiments, the kit may include multiple containers, for example, a first container that includes one or more isolated nucleic acids, peptides, and / or proteins disclosed herein.

[0075] Nucleic acids, cloning and expression systems The present disclosure further provides artificial nucleic acid molecules. Nucleic acids may include DNA or RNA, and may be wholly or partially synthetic or recombinant. Reference to a nucleotide sequence described herein includes DNA molecules having the specified sequence, and, unless the context requires otherwise, includes RNA molecules having the specified sequence in which U is replaced with T or its derivatives, such as pseudouridine. Other nucleotide derivatives or modified nucleotides may be incorporated into the artificial nucleic acid molecule.

[0076] The present disclosure also provides constructs in the form of vectors (e.g., plasmids, phagemids, cosmids, transcription or expression cassettes, artificial chromosomes, etc.) that contain artificial nucleic acid molecules encoding nucleic acid sequences generated as disclosed herein. The present disclosure further provides host cells that contain one or more constructs as described above.

[0077] Also provided are methods for producing isolated peptides and / or proteins using recombinant techniques known in the art and described above. Recombinant protein production and expression are well known in the art and can be performed using conventional procedures such as those disclosed in Sambrook et al., Molecular Cloning: A Laboratory Manual (4th Ed. 2012), Cold Spring Harbor Press. For example, expression of HA or NA polypeptides can be achieved by culturing host cells containing artificial nucleic acid molecules encoding HA or NA as disclosed herein under appropriate conditions. For example, expression of recombinant HA or NA polypeptides can be achieved by culturing host cells containing nucleic acid molecules encoding HA or NA as disclosed herein under appropriate conditions. After production by expression, HA or NA can be isolated and / or purified using any suitable technique and then used as appropriate.

[0078] Systems for cloning and polypeptide expression in a variety of different host cells are well known in the art. Any protein expression system (e.g., stable or transient) compatible with the constructs disclosed in this application may be used to produce the generated amino acid sequences described herein.

[0079] Suitable vectors can be selected or constructed to contain appropriate regulatory sequences, such as promoter sequences, terminator sequences, polyadenylation sequences, enhancer sequences, marker genes, and other sequences as appropriate.

[0080] To express the generated amino acid sequence as disclosed herein, the nucleic acid encoding the generated amino acid sequence can be introduced into a host cell. The introduction can use any available technique. For eukaryotic cells, suitable techniques can include calcium phosphate gene transfer, DEAE-dextran, electroporation, liposome-mediated gene transfer, and transduction using retroviruses or other viruses, such as vaccinia, or baculovirus in the case of insect cells. For bacterial cells, suitable techniques can include calcium chloride transformation, electroporation, and gene transfer using bacteriophage. These techniques are well known in the art. (See, for example, "Current Protocols in Molecular Biology", Ausubel et al. eds., John Wiley & Sons, 2010). DNA transfer can be followed by a selection method (e.g., antibiotic resistance) to select cells containing the vector.

[0081] The host cell may be a plant cell, a yeast cell, or an animal cell. Animal cells include invertebrate cells (e.g., insect cells), non-mammalian vertebrate cells (e.g., birds, reptiles, and amphibians), and mammalian cells. In one embodiment, the host cell is a mammalian cell. Examples of mammalian cells include, but are not limited to, COS-7 cells, HEK293 cells; baby hamster kidney (BHK) cells; Chinese hamster ovary (CHO) cells; mouse Sertoli cells; African green monkey kidney cells (VERO-76); human cervical carcinoma cells (e.g., HeLa); canine kidney cells (e.g., MDCK), and the like. In one embodiment, the host cell is a CHO cell. In one embodiment, the host cell is an insect cell.

[0082] How to use The present disclosure provides a method of administering a vaccine composition described herein to a subject. The method may be used to vaccinate a subject against a virus (e.g., influenza virus). In some embodiments, the vaccination method comprises administering to a subject in need of vaccination a vaccine composition comprising one or more isolated nucleic acids, peptides, and / or proteins encoding the generated amino acid sequences as described herein (e.g., a recombinant influenza virus HA described herein or a recombinant influenza virus NA described herein) and an optional adjuvant in an amount effective to vaccinate the subject against the virus (e.g., influenza virus). Similarly, the present disclosure provides a vaccine composition comprising one or more isolated nucleic acids, peptides, and / or proteins encoding the generated amino acid sequences as described herein (e.g., influenza virus HA or NA) and an optional adjuvant for use in vaccinating a subject against a virus (e.g., influenza virus) (or for the manufacture of a medicament for use in vaccination).

[0083] The present disclosure also provides a method of immunizing a subject against a virus (e.g., an influenza virus), comprising administering to the subject an immunologically effective amount of a vaccine composition comprising one or more recombinant influenza virus HA or NA described herein, and an optional adjuvant.

[0084] In some embodiments, the method or use prevents infection or disease by a virus (e.g., influenza virus) in a subject. In some embodiments, the method or use induces a protective immune response in the subject. In some embodiments, the protective immune response is an antibody response.

[0085] The immunization methods / uses provided herein may induce a broad neutralizing immune response against one or more viruses (e.g., influenza viruses). Thus, in various embodiments, the compositions described herein may provide broad cross-protection against various types of viruses (e.g., influenza viruses). In some embodiments, the compositions provide cross-protection against avian, swine, seasonal and / or pandemic influenza viruses. In some embodiments, the immunization methods / uses may induce an improved immune response against one or more seasonal influenza strains (e.g., standard of care strains). For example, the improved immune response may be an improved humoral immune response. In some embodiments, the immunization methods / uses may induce an improved immune response against one or more pandemic influenza strains. In some embodiments, the immunization methods / uses may induce an improved immune response against one or more swine influenza strains. In some embodiments, the immunization methods / uses may induce an improved immune response against one or more avian influenza strains.

[0086] In certain embodiments, provided herein are methods of enhancing or expanding a protective immune response, the methods comprising administering to a subject an immunologically effective amount of a vaccine composition disclosed herein, where the vaccine composition increases the vaccine efficacy of a standard of care influenza virus vaccine composition by an amount ranging from about 5% to about 100%, e.g., about 10% to about 25%, about 20% to about 100%, about 15% to about 75%, about 15% to about 50%, about 20% to about 75%, about 20% to about 50%, or about 40% to about 80%, e.g., about 40% to about 60% or about 60% to about 80%. In certain embodiments, the vaccine compositions disclosed herein have a vaccine efficacy that is at least 5% greater than the vaccine efficacy of a standard of care influenza virus vaccine, e.g., at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, or at least 100% greater than the vaccine efficacy of a standard of care influenza virus vaccine. Similarly, the disclosure provides any of the vaccine compositions described herein for use (or for the manufacture of a medicament for use in enhancing or expanding a protective immune response in a subject).

[0087] Also provided are methods of preventing a viral disease (e.g., influenza virus disease) in a subject, comprising administering a vaccine composition comprising one or more isolated nucleic acids encoding generated amino acid sequences, peptides, and / or proteins (e.g., a recombinant influenza virus HA or NA as described herein), and an optional adjuvant, in an amount effective to prevent the viral disease (e.g., influenza virus disease) in the subject. Similarly, the disclosure provides vaccine compositions comprising one or more recombinant influenza virus HA or NA, and an optional adjuvant, for use (or for the manufacture of a medicament for use therein) in preventing a viral disease (e.g., influenza virus disease) in a subject.

[0088] Also provided is a method of inducing an immune response to influenza virus HA and influenza virus NA in a subject, comprising administering to the subject a vaccine composition comprising one or more recombinant influenza virus HAs described herein, one or more recombinant influenza virus NAs described herein, and an optional adjuvant.

[0089] 1 is a block diagram of an example system 100 that can be used to manufacture vaccines. In the system 100, new vaccines 116 are designed and manufactured using the techniques described herein. For example, for viruses with many strains and / or rapidly mutating strains, such as influenza, human rhinovirus, HIV, coronavirus viruses such as coronavirus disease 2019, or new viruses never before encountered, the techniques described herein can be used to rapidly generate vaccine candidates that can be tested for use in humans or other subjects.

[0090] As inputs, the system 100 receives virus strain data 102 and wild-type amino acid data 104. The virus strain data 102 includes data regarding one or more virus strains for which a vaccine is desired. This virus strain data 102 may include amino acid sequence data as well as other types of data, such as metadata (e.g., unique identifiers, strain identification) or non-metadata characteristics (e.g., records of physicochemical properties of the amino acid sequence, such as molecular weight). The wild-type amino acid data 104 may include a corpus of amino acid definitions for hundreds, thousands, hundreds of thousands, or more amino acid sequences. These sequences are referred to herein as wild-type, which in some embodiments indicates that they are amino acid sequences that are commonly found in the wild. However, in other embodiments, the amino acid sequences may include previously unseen artificial amino acid sequences, sequences of human-manufactured amino acids, or other types of amino acids. The amino acid data 104 may include amino acid sequence data as well as other types of data, such as metadata (e.g., unique identifiers, strain identification) or non-metadata characteristics (e.g., records of physicochemical properties of the amino acid sequence, such as molecular weight).

[0091] The system 100 includes a computer system 106 that can use the data 102 and 104 to generate data 108 of candidate non-wild-type amino acid sequences. These non-wild-type amino acid sequences are amino acids that are not found in the wild or are not known to be found in the wild. As will be appreciated, one or more of the candidate non-wild-type amino acid sequences 108 may in fact exist in the wild, but may be unknown to the operators of the system 100, or even to the community at large. The candidate non-wild-type amino acid data 108 may include amino acid sequence data as well as other types of data, such as metadata (e.g., unique identifiers, strain identification) or non-metadata characteristics (e.g., records of physicochemical properties of the amino acid sequence, such as molecular weight).

[0092] The computer system 106 validates one or more of the candidates in the data 108 for production, resulting in data 110. The data 110 may include amino acid sequence data as well as other types of data, such as metadata (e.g., unique identifiers, strain identification) or non-metadata characteristics (e.g., records of physicochemical properties of the amino acid sequence, such as molecular weight). In some cases, the data 102 / 104, 108, and 110 are in the same data format, and in some cases, the data 102 / 104, 108, and 110 are in different data formats.

[0093] In some cases, the validation process used to select candidates may include determining whether the amino acid sequence can be synthesized at all, or whether it can be easily or economically synthesized. As will be appreciated, an amino acid sequence may define the structure of such a molecule that cannot exist in the physical world due to the geometry and forces that the molecule exhibits. Thus, such impossible sequences may be excluded from the validation process. In addition, some of the candidates may be excluded despite defining effective molecules. For example, the computing system 106 may maintain a data store of previous candidates that were not actually effective as a vaccine once investigated in clinical trials, or predicted to be less immunogenic or less protective against the virus strain of interest, which may include the virus strain data 102. In such cases, the candidates in the data 108 may be excluded from the validated data 110. In some cases, candidates may be excluded or prioritized based on synthesis and manufacturing considerations. For example, candidates with particular synthesis or processing conditions (e.g., cold storage, shock sensitivity) may be excluded from validation or de-prioritized compared to other candidates with less demanding synthesis or processing conditions.

[0094] The system 100 may also include a vaccine production device 112 capable of using the vaccine precursor 114 and one or more verified non-wild type amino acid sequence data 110 to produce one or more vaccine doses or vaccine molecules 116. As will be appreciated, early exploration and testing will require much smaller scale synthesis than large scale production of a vaccine that has been tested, found to be safe and effective, and approved for use in humans or other subjects. Thus, the details of the production device 112 may vary according to needs. Similarly, the vaccine precursor 114 includes the items, chemicals, materials, etc. for the production of the vaccine 116, but the precursor 114 will likewise vary according to needs.

[0095] Figure 2 is a schematic diagram of data that can be used to manufacture a vaccine. For example, the data shown here can be used by computer system 106 or other computer systems. Broadly speaking, data 104 is transformed into a lower dimensional space and modified to create new amino acid sequences, one or more of which are then selected for use in vaccine manufacture. This data can be used by computer system 106 or other computing systems.

[0096] Wild-type amino acid sequence 104 is one or more data objects that define a plurality of wild-type amino acid sequences. Wild-type amino acid data 104 is shown here with subsections of the sequence rendered for readability using single-letter designations recommended by the International Union of Pure and Applied Chemistry-International Union of Biochemistry and Molecular Biology (IUPAC-IUBMB) Joint Commission on Biochemical Nomenclature. For wild-type amino acid sequences, data 104 may include vectors of data values ​​(e.g., single American Standard Code for Information Interchange (ASCII) characters, integers) to represent amino acids in the sequence represented by data 104. As will be appreciated, longer sequences may have more indexes than are visually shown here, and more sequences may be stored in data 104 than are shown. Additionally, for clarity, other portions of data 104 are not shown here. Each amino acid sequence may be recorded as either a single character or a string of characters. A string of characters may include multiple single characters. The one or more amino acid sequences may include a first amino acid sequence and a second amino acid sequence, each of the first amino acid sequence and the second amino acid sequence including a respective single character or a respective string of characters. That is, each amino acid sequence can be stored with different values ​​yet conform to the same format, allowing for interoperability and consistent processing of the data.

[0097] As will be appreciated, vectors in data 104 have a length, which may all be the same between vectors. These vectors define a particular dimension of data 104. For example, a length of 632 defines a space with 632 dimensions, a length of 88 defines a space with 88 dimensions, etc. For a sequence that can contain one of 20 different amino acids, each dimension has 20 domains or sizes. Thus, the corpus of amino acid sequences in data 104 defines a distribution of vectors (or point positions) in a space with a dimension of the length of the amino acid sequence.

[0098] Data 104 undergoes variational encoding (described elsewhere) to generate multiple reduced-dimension arrays 202 in a reduced-dimension space from one or more data objects. In this example, the reduced-dimension space has five dimensions, and data 202 may be stored in a vector of length 5, although different dimensions (and therefore different vector lengths) may be used in other implementations. Data 202 may store data (e.g., real numbers) or other suitable data at each index of the vector, where a value is encoded from a value in the amino acid sequence of data 104 with the variational data appended from the variational encoding. In some implementations, these real numbers are trained so that 1) similar sequences are proximate and 2) the decoder portion of the model can be used to decode the numeric coordinates. Thus, each reduced-dimension array includes data for at least one respective wild-type amino acid sequence.

[0099] In some examples, this variational encoding involves a lossy data transformation, resulting in data 202 that is based on, but does not contain, all of the information contained in data 104. However, this is nevertheless an advantageous process because it can enable the operations described herein to generate novel, non-wild-type amino acid sequences useful in vaccine development and production.

[0100] The reduced dimensional space is of lower dimension than the corresponding wild-type amino acid sequence. This may allow computational operations in the lower dimensional space that would be impossible, computationally inefficient, or otherwise undesirable in the high dimensional space of the data 104. For example, the dimensions of the reduced dimensional space are not collinear with or represent a small subset of the dimensions of the high dimensional space, so that single dimensional properties do not map directly to a single dimension of the high dimensional space. Thus, operations in a single dimension of the reduced dimensional space may allow efficient execution and produce results that are impossible or unintuitive when operating or thinking in a higher dimensional space.

[0101] In some implementations, the data 206 stores candidate sequences generated from the data 202. One such example of random sampling is performed from the entire normal distribution space defined by the data 202. For example, if the data 202 has five dimensions, then typically, selections are made across the five axes.

[0102] In some implementations, data 206 stores candidate sequences generated from intervening data 204. One such example is the assembly of distributions in each dimension of a reduced dimensional space. Because data 202 is stored as a plurality of vectors, a statistical distribution of values ​​in a particular dimension can be assembled in data 204. For example, if integers are stored at each index of a vector in data 202, then each index of a vector in data 204 can store a histogram of integers at the same index in a vector in data 202. In another example, if real numbers are stored at each index of a vector in data 202, then each index of a vector of data 204 can store parameters of a function that defines a best fit curve as found by linear regression or similar analysis. As will be appreciated, the type of data stored at each index of data 204 can be determined based on the type of data stored at each index of data 202. In this manner, a plurality of reduced dimensional arrays define a distribution of values ​​along each dimension of a reduced dimensional space.

[0103] Data 206 stores candidate sequences. For example, multiple candidate sequences can be generated in the same reduced dimensional space used for data 202 and 204. This can be done many times (tens, hundreds, thousands, tens of thousands, hundreds of thousands, millions, billions, trillions, or more) to create many candidate sequences. Data 206 can store definitions of amino acid sequences that have similar properties to those in data 104 (although stored in a lower dimensional space like data 202), but that may not actually appear in data 104. If a sequence in data 104 has particularly beneficial or desirable properties, these properties can be expected to be found in at least some of data 206. For example, if an amino acid sequence in data 104 elicits an immune response in a subject, sequences defined by data 206 are likely to provide a similar immune response. Also, they are likely to elicit a greater or lesser immune response because they differ to some extent from the amino acids in sequence 104. So, as will be explained, this can generate new sequences not previously known or previously understood, eliciting greater immune responses and making them suitable for use in vaccines, thus beneficially advancing vaccine technology.

[0104] In some cases, data 204 is randomly sampled according to its distribution to generate data 206. For example, if each index includes a histogram, values ​​in the histogram are selected weighted by the height of each value in the histogram. For example, if each index includes parameters of a curve that defines a Y value given an X value, the X value may be weighted by the height of the curve or selected by randomly selecting one of the points under the curve. In this way, the distribution values ​​at a given index in data 202 will be similar to the distribution of values ​​at the same index in data 206, but they are unlikely to be statistically identical.

[0105] As will be described, each of the candidate sequences in the data 206 can then be tested and the best candidate selected for use in analysis or vaccine production. For example, an immune response predictor, such as an antibody titer predictor, can be used to predict an immune response in a subject for a given viral amino acid sequence. The titer predictor can be configured to accept two amino acid sequences as input. The function can be configured to return a predicted immune response of the subject (e.g., human, animal) as an output. The output can take the form of, for example, a value between 0 and 1, with higher values ​​indicating a predicted higher immune response. The prediction function can operate using a machine learning model.

[0106] To perform this operation, data 102 containing viral amino acid sequences is modified in the same manner as data 104, resulting in the formation of data 208 of the same format and type as data 202. That is, data 102 holding one or more data objects defining viral amino acid sequences undergoes variational encoding (described elsewhere) to generate reduced-dimensional viral sequence 208. This data 208 stores the data in the same reduced-dimensional space as data 202-206, allowing efficient computational operations on any of data 202-208.

[0107] For each sequence in the data 206, the potency predictor generates a candidate score. The candidate score is the predicted immune response for the amino acid sequence. Three examples of many sequences and scores are shown here, but it is understood that the potency predictor can be used many times (tens, hundreds, thousands, tens of thousands, hundreds of thousands, millions, billions, trillions, or more) to create as many candidate scores as there are candidate sequences.

[0108] These candidate scores indicate the expected immune response level and can therefore be considered a prediction of how effective the candidate sequence is likely to be in a vaccine. At least one selected candidate sequence is selected from the candidate sequences. Various computational processes can be used to identify the "best" sequence for testing and / or production.

[0109] In one example, a predefined or dynamically defined number of candidate sequences are selected. This includes selecting the candidate sequence with the N highest candidate scores. The value of N can be based on, for example, the throughput of the device and system on which the vaccine can be tested, so if N amino acid sequences can be tested, the same N value can be used here. If N is a value greater than 1, multiple candidate sequences are selected. If N is a value of 1, a single candidate sequence is selected. Shown here is an example where N has a value of 2, such that 2 sequences are selected.

[0110] In one example, a predefined or dynamically defined threshold is used to select candidate sequences. The threshold may be based, for example, on a minimum value expected to produce good results. As will be appreciated, the threshold may be close to a maximum value (e.g., close to but less than 1) to select only the most promising candidate sequences, close to a minimum value (e.g., close to but greater than 0) to select all but the least promising candidates, or any other suitable value. This may result in candidate sequences not being selected in some cases, depending on the threshold and the quality of the candidates.

[0111] Data 110 is created by constructing amino acid sequences in the high-dimensional space used in data 102 and 104. Depending on the configuration of the operations, it may be possible that a single representation in data 212 can be mapped to more than one amino acid sequence. As explained above, the conversion from a high-dimensional space to a low-dimensional space can be lossy. In such cases, this may mean that any given sequence representation in the low-dimensional space may be ambiguous and may specify more than one actual amino acid. In the example shown, one candidate sequence is used to create one new amino acid sequence, while another candidate sequence is used to create two new amino acid sequences, but more than two may be possible.

[0112] The described data processing constraints allow novel amino acid sequences in data 110 to retain a certain amount of similarity (e.g., as defined by edit distance or other metric) with wild-type amino acid sequences in data 104. For clarity, differences between data 104 and 110 are represented by bolding certain letters in data 110.

[0113] It will be understood that one of the new amino acid sequences in data 110 can, but need not, be the same as one of the wild-type amino acid sequences in data 104. Additionally, one of the new amino acid sequences 110 can, but need not, be the same as a wild-type amino acid sequence found in nature that is not involved in the data manipulation described herein. Additionally, one of the new amino acid sequences 110 can, but need not, be the same as another new amino acid sequence that was previously created in this or another data processing operation, tested for potential as a vaccine, and discarded (e.g., due to low efficacy, safety concerns, inmanufacturability, or other undesirable properties). Thus, in some cases, data 110 may be filtered to remove known new amino acid sequences, leaving only unknown or uncharacterized amino acid sequences.

[0114] Figure 3 is a flow chart of an example process 300 that may be used to process high dimensional data in a lower dimensional space, such as may be used in the manufacture of a vaccine. For example, process 300 may be performed using the data shown in Figures 1 and 2, e.g., 102 / 104, 110, 202-212, and therefore elements of those figures will be used for illustration. Possible embodiments of various elements of process 300 are described below in processes 400-700.

[0115] One or more data objects defining a plurality of wild-type amino acid sequences are received 302. For example, computer system 106 may access data 104 from a disk, receive data 104 over a network connection, etc.

[0116] A plurality of reduced dimensional arrays are generated in the reduced dimensional space (304). For example, the computer system 106 may use one or more data processing operations that use the data 104 as input and generate the data 202 as output. In doing so, the computer system 106 may embed variability in the data 202. An example of 304 is described in more detail in process 400.

[0117] One or more data objects defining viral amino acid sequences are received 306. For example, computer system 106 may access data 102 from a disk, receive data 104 over a network connection, etc.

[0118] At least one reduced-dimensional viral sequence in the reduced-dimensional space is generated (308). For example, the computer system 106 may use one or more data processing operations that use the data 102 as input and generate the data 208 as output. In some examples, the computer system 106 may embed variability into the data 208 in the same manner as performed in 304 (see, e.g., process 400). In other examples, the computing system determines whether to embed the variability in a different manner or not at all.

[0119] A plurality of candidate sequences in the reduced dimensional space are generated using the plurality of reduced dimensional arrays (310). For example, the computer system 106 may analyze the data 202 to generate the data 204. To do so, the computer system 106 may characterize values ​​of various vectors of the data 202 and record such characterizations in the data 204. In some cases, the computer system 106 creates the plurality of candidate sequences in the reduced dimensional space by sampling a distribution of values ​​of the plurality of reduced dimensional arrays.

[0120] Each candidate sequence is scored to generate a candidate score (312). For example, computer system 106 can analyze data 206 and 208 to generate data 210. In some cases, computer system 106 can use a predictor or classifier trained on historical data of amino acids in a low-dimensional space to generate a prediction of a biological response (e.g., the strength at which a subject will produce antibodies). An example of 312 is described in more detail in process 500.

[0121] At least one candidate sequence is selected as a selected candidate sequence (314). For example, the computing system 106 can generate the data 212 from the data 210. The computing system 106 can select the selected candidate sequence using, for example, the candidate scores. An example of 314 is described in more detail in process 600.

[0122] For each selected candidate sequence, at least one new amino acid sequence is generated (316). For example, computing system 106 can generate data 110 from data 212 and provide data 110 for use in vaccine production. To do this, computing system 106 can find points or vectors in high-dimensional space that correspond to points or vectors in data 212. As will be appreciated, the projection of vectors from low-dimensional space to high-dimensional space can define a domain of results instead of a single point result. Thus, in some cases, computer system 106 can generate all valid amino acid sequences that are in the domain of results, resulting in two or more new amino acid sequences in data 110 for each vector in data 212.

[0123] A vaccine for each new amino acid sequence is produced (318). The vaccine may include a protein and / or a nucleic acid defined by the new amino acid sequence, or any other delivery vehicle, including a viral or bacterial vector, whereby such a nucleic acid or delivery vehicle produces a protein defined by the new amino acid sequence (318). For example, the computing device 106 and / or the vaccine production device 112 may operate to use the data 110 and the vaccine precursor 114 to create a vaccine 116. This production may be small batches for initial testing, clinical trials, and / or public use. As will be appreciated, elements 316 and 318 may be separated by a significant amount of time and intervening operations. For example, if the production at 318 is a large-scale production for public use, this may occur only after clinical trials have demonstrated that the vaccine is safe and effective for its intended purpose.

[0124] 4 is a flow chart of an example process 400 that may be used to process high dimensional data in a low dimensional space that may be used to manufacture a vaccine, including creating a representation of a wild type amino acid sequence using a variational autoencoder that predicts the mean and variance of the input data. For example, process 400 may be performed using the data shown in Figures 1 and 2, and therefore the description will use elements of those figures. Process 400 is one possible example of how operation 304 may be performed, although other processes may be used.

[0125] One or more variational autoencoders, which are described below, are accessed 402. For example, the computer system 106 may access the autoencoder's data from a disk, receive the data over a network connection, etc. This data may define one or more functions, libraries, modules, etc. that operate on input data and return output data.

[0126] The variational autoencoder creates a low-dimensional representation of the amino acid sequence 404. For example, the computing system 106 can run a variational autoencoder on the data 104 to generate the data 202.

[0127] The reduced dimensional array is received 406. For example, the computing system 106 can receive the data 202 from the variational autoencoder, which may include accessing the data from a disk, receiving the data 202 over a network connection, etc.

[0128] 5 is a flow chart of an example process 500 that may be used to process high dimensional data in a lower dimensional space, such as may be used in the manufacture of a vaccine. For example, process 500 may be performed using the data shown in Figures 1 and 2, and therefore the description will use elements of those figures. Process 500 is one possible example of how operation 312 may be performed, although other processes may be used.

[0129] Each of the candidate sequences and the reduced-dimension viral sequence are provided as inputs to the antibody titer predictor (502). For example, the computer system 106 may access the titer predictor data from a disk, receive the data over a network connection, etc. This data may define one or more functions, libraries, modules, etc. that operate on the input data and return output data. This operation may be performed sequentially on one or more reduced-dimension viral sequences.

[0130] A prediction is generated using a titer predictor 504. For example, the computing system 106 can execute the titer predictor using the data 206 and the data 208 to produce the data 210.

[0131] The candidate score for each candidate sequence is received as output from the potency predictor 506. For example, the computing system 106 can receive the data 210 from the potency predictor, which may include accessing the data from a disk, receiving the data 202 over a network connection, etc.

[0132] 6 is a flow chart of an example process 600 that may be used to process high dimensional data in a lower dimensional space, such as may be used in the manufacture of a vaccine. For example, process 600 may be performed using the data shown in Figures 1 and 2, and therefore the description will use elements of those figures. Process 600 is one possible example of how operation 314 may be performed, although other processes may be used.

[0133] The candidate sequences are sorted by candidate score 602. For example, the computer system 106 can order the data 210 into a list in memory such that each entry in the list has a candidate score greater than or equal to (or less than or equal to) the subsequent entry.

[0134] The top candidate scores are identified (604). For example, the computer system 106 may identify some of the candidate sequence pairs:candidate scores from the top (or bottom) of the list. In some cases, the computer system 106 selects the N pairs with the highest candidate scores, where N is some positive integer value. In some cases, when the threshold is less than the maximum possible candidate score and greater than the minimum possible candidate score, the computer 106 selects all pairs with candidate scores greater than the threshold.

[0135] Candidate sequences corresponding to the top candidate scores are selected 606. For example, the computer system 106 can select candidate sequences among the sequence:candidate score pairs identified.

[0136] 7 is a swim lane diagram of a process 300 for manufacturing a vaccine. To execute the elements of the process 300, the computer system 106 may use operational elements such as a data handler 702, a variational autoencoder 704, and an immune response predictor 706, although other computational architectures may be used. Each element 702-706 may be implemented as one or more programs, routines, libraries, modules, etc. that execute on the computer system 106 and that may communicate, store, and manipulate data, such as the data shown in FIGS. 1 and 2. As will be appreciated, the various elements 702-706 may be executed in hardware separate from the hardware executing other elements of the computing system 106.

[0137] A data handler 702 runs on the computing system 106 and handles data operations such as accessing data on disk, transferring data over a network connection within the computing system 106, and manipulates data such as 302, 310, 314, and 316 in addition to other operations.

[0138] The variational autoencoder 704 includes one or more computational models, such as a linear support vector machine (linear SVM), boosting of other algorithms (e.g., AdaBoust), neural networks, logistic regression, naive Bayes, memory-based learning, random forests, bag trees, decision trees, boosted trees, or boosted stumps. These models can operate on input data of a given dimension and generate output data for each of the lower dimensional input data. The variational autoencoder 704 can compress the input information into a constrained multivariate latent distribution and can also reconstruct such data into the format of the input. Some implementations of the variational autoencoder can operate on input data characterized by an unknown probability distribution and approximate the distribution of that data. Intervening operations of the encoding and reconstruction functions include, but are not limited to, predicting the mean and variance of the input data.

[0139] The titer predictor 706 includes one or more computational models such as a linear support vector machine (linear SVM), boosting of other algorithms (e.g., AdaBoost), neural networks, logistic regression, naive Bayes, memory-based learning, random forests, bug trees, decision trees, boosted trees, or boosted stumps, etc. These models can be trained based on a dataset of sequences in a low-dimensional space tagged with outcome values ​​indicative of a biological response, such as antibody titer, when the sequences are introduced into a subject (e.g., a human, mammal, patient).

[0140] 8 illustrates an example of a computing device 800 and an example of a mobile computing device that can be used to implement the techniques described herein. The computing device 800 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. The mobile computing device is intended to represent various forms of mobile devices, such as personal digital assistants, mobile phones, smartphones, and other similar computing devices. The components, their connections and relationships, and their functions illustrated herein are intended to be merely exemplary and are not intended to limit the embodiments of the invention described and / or claimed herein.

[0141] The computing device 800 includes a processor 802, a memory 804, a storage device 806, a high-speed interface 808 connecting to the memory 804 and multiple high-speed expansion ports 810, and a low-speed interface 812 connecting to a low-speed expansion port 814 and the storage device 806. Each of the processor 802, the memory 804, the storage device 806, the high-speed interface 808, the high-speed expansion port 810, and the low-speed interface 812 are interconnected using various buses and may be mounted on a common motherboard or otherwise as appropriate. The processor 802 may process instructions, including instructions stored in the memory 804 or the storage device 806, for execution within the computing device 800 to display graphical information in a GUI on an external input / output device, such as a display 816 coupled to the high-speed interface 808. In other implementations, multiple processors and / or multiple buses may be used along with multiple memories and memory types as appropriate. Also, multiple computing devices may be connected, with each device providing a portion of the required operations (e.g., as a server bank, a group of blade servers, or a multiprocessor system).

[0142] The memory 804 stores information within the computing device 800. In some implementations, the memory 804 is one or more volatile memory units. In some implementations, the memory 804 is one or more non-volatile memory units. The memory 804 may also be another form of computer-readable medium, such as a magnetic disk or optical disk.

[0143] The storage device 806 can provide mass storage for the computing device 800. In some implementations, the storage device 806 can be or include a computer-readable medium such as a floppy disk device, a hard disk device, an optical disk device, or an array of devices including a tape device, a flash memory or other similar solid-state memory device, or a device in a storage area network or other configuration. The computer program product can be tangibly embodied in an information carrier. The computer program product can also include instructions that, when executed, perform one or more methods, such as those described above. The computer program product can also be tangibly embodied in a computer or machine-readable medium, such as the memory 804, the storage device 806, or a memory on the processor 802.

[0144] The high-speed interface 808 manages the bandwidth-intensive operations of the computing device 800, and the low-speed interface 812 manages the less bandwidth-intensive operations. Such an allocation of functions is merely exemplary. In some implementations, the high-speed interface 808 is coupled to the memory 804, the display 816 (e.g., via a graphics processor or accelerator), and a high-speed expansion port 810 that can accept various expansion cards (not shown). In this implementation, the low-speed interface 812 is coupled to the storage device 806 and the low-speed expansion port 814. The low-speed expansion port 814 can be coupled to one or more input / output devices, such as a keyboard, a pointing device, a scanner, or a networking device, such as a switch or router, via, for example, a network adapter, which may include various communication ports (e.g., USB, Bluetooth, Ethernet, wireless Ethernet).

[0145] Computing device 800 may be implemented in a number of different forms, as shown. For example, it may be implemented as a standard server 820 or multiple times in a group of such servers. It may also be implemented in a personal computer, such as a laptop computer 822. It may also be implemented as part of a rack server system 824. Alternatively, components from computing device 800 may be combined with other components in a mobile device (not shown), such as mobile computing device 850. Each such device may include one or more of computing device 800 and mobile computing device 850, and the entire system may be composed of multiple computing devices in communication with each other.

[0146] The mobile computing device 850 includes a processor 852, a memory 864, input / output devices such as a display 854, a communication interface 866, and a transceiver 868, among other components. The mobile computing device 850 may also be provided with a storage device, such as a microdrive or other device, to provide additional storage. Each of the processor 852, memory 864, display 854, communication interface 866, and transceiver 868 are interconnected using various buses, and some of the components may be mounted on a common motherboard or in other suitable manner.

[0147] The processor 852 can execute instructions within the mobile computing device 850, including instructions stored in the memory 864. The processor 852 can be implemented as a chipset of chips including separate analog and digital processors. The processor 852 can provide for coordination of other components of the mobile computing device 850, such as control of a user interface, applications run by the mobile computing device 850, and wireless communication by the mobile computing device 850.

[0148] The processor 852 can communicate with a user via a control interface 858 and a display interface 856 coupled to a display 854. The display 854 can be, for example, a TFT (thin film transistor liquid crystal display) display, an OLED (organic light emitting diode) display, or other suitable display technology. The display interface 856 can include appropriate circuitry for driving the display 854 to present graphical and other information to the user. The control interface 858 can receive commands from the user and translate them for submission to the processor 852. Additionally, an external interface 862 can provide communication with the processor 852 to enable short-range area communication between the mobile computing device 850 and other devices. The external interface 862 can be provided, for example, for wired communication in some implementations or for wireless communication in other implementations, and multiple interfaces can also be used.

[0149] The memory 864 stores information within the mobile computing device 850. The memory 864 may be implemented as one or more of one or more computer readable media or media, one or more volatile memory units, or one or more non-volatile memory units. Expansion memory 874 may also be provided and connected to the mobile computing device 850 via an expansion interface 872, which may include, for example, a Single In Line Memory Module (SIMM) card interface. The expansion memory 874 may provide additional storage space for the mobile computing device 850 or may store applications and other information for the mobile computing device 850. In particular, the expansion memory 874 may include instructions for performing or supplementing the above processes, and may also include secure information. Thus, for example, the expansion memory 874 may be provided as a security module for the mobile computing device 850 and may be programmed with instructions that enable secure use of the mobile computing device 850. Additionally, secure applications may be provided via the SIMM card, such as placing identifying information on the SIMM card in an unhackable manner, along with additional information.

[0150] The memory may include, for example, flash memory and / or NVRAM memory (non-volatile random access memory), as described below. In some implementations, the computer program product is tangibly embodied in an information carrier. The computer program product includes instructions that, when executed, perform one or more methods, such as those described above. The computer program product may be a computer-readable or machine-readable medium, such as memory 864, expansion memory 874, or memory on processor 852. In some implementations, the computer program product may be received in a propagated signal, for example, via transceiver 868 or external interface 862.

[0151] The mobile computing device 850 can communicate wirelessly via a communication interface 866, which may include digital signal processing circuitry as needed. The communication interface 866 can provide for communication under various modes or protocols, such as GSM voice (Global System for Mobile Communications), SMS (Short Message Service), EMS (Enhanced Message Service), or MMS messaging (Multimedia Messaging Service), CDMA (Code Division Multiple Access), TDMA (Time Division Multiple Access), PDC (Personal Digital Cellular), WCDMA (Wideband Code Division Multiple Access), CDMA2000, or GPRS (General Packet Radio Service). Such communication can be via a transceiver 868, for example, using radio frequencies. In addition, short-range communication can be performed, such as using Bluetooth, WiFi, or other such transceivers (not shown). In addition, a GPS (Global Positioning System) receiver module 870 can provide additional navigation and location related wireless data to the mobile computing device 850, which can be used as appropriate by applications executing on the mobile computing device 850.

[0152] The mobile computing device 850 can also communicate voice using an audio codec 860 that can receive voice information from a user and convert it into usable digital information. The audio codec 860 can also generate audible sounds for the user, for example, through a speaker in a handset of the mobile computing device 850. Such sounds can include sounds from a voice call, can include recorded sounds (e.g., voice messages, music files, etc.), and can include sounds generated by applications running on the mobile computing device 850.

[0153] The mobile computing device 850 can be implemented in a number of different forms, as shown in the figure, For example, it can be implemented as a mobile phone 880. It can also be implemented as part of a smartphone 882, personal digital assistant, or other similar mobile device.

[0154] Various implementations of the systems and methods described herein may be realized in digital electronic circuitry, integrated circuitry, specially designed ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementation in one or more computer programs executable and / or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, coupled to transmit and receive data and instructions from a memory device, at least one input device, and at least one output device.

[0155] These computer programs (also referred to as programs, software, software applications or code) include machine instructions for a programmable processor and may be implemented in high-level procedural and / or object-oriented programming languages ​​and / or assembly / machine languages. As used herein, the terms machine-readable medium and computer-readable medium refer to any computer program product, apparatus, and / or device (e.g., magnetic disk, optical disk, memory, programmable logic device (PLD)) used to provide machine instructions and / or data to a programmable processor, including a machine-readable medium that receives the machine instructions as a machine-readable signal. The term machine-readable signal refers to any signal used to provide machine instructions and / or data to a programmable processor.

[0156] To provide for user interaction, the systems and techniques described herein can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user, and a keyboard and pointing device (e.g., a mouse or trackball) by which the user can provide input to the computer. Other types of devices may be used to provide for user interaction as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback), and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0157] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), includes middleware components (e.g., an application server), or includes front-end components (e.g., a client computer having a graphical user interface or a web browser that allows a user to interact with an implementation of the systems and techniques described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0158] A computing system may include clients and servers. Clients and servers are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

Claims

1. 1. A dimensionality reduction method for generating amino acid sequences, carried out by one or more computer systems, comprising: receiving one or more data objects defining a plurality of wild type amino acid sequences; generating a plurality of reduced-dimensional arrays in a reduced-dimensional space from the one or more data objects, each reduced dimension array comprises data for at least one respective one of said wild-type amino acid sequences; the reduced dimensional space is less dimensional than the wild-type amino acid sequence; the plurality of reduced dimensional arrays defining a distribution of values ​​along each dimension of the reduced dimensional space; Generating and generating a plurality of candidate sequences in the reduced dimensional space using the plurality of reduced dimensional sequences; receiving one or more data objects defining a viral amino acid sequence; generating at least one reduced-dimensionality viral sequence within the reduced-dimensionality space; providing each of said candidate sequences and at least one of said reduced dimensionality viral sequences as inputs to a titer predictor; receiving a candidate score for each of the candidate sequences as output from a potency predictor; selecting at least one candidate sequence from among the candidate sequences; generating at least one new amino acid sequence for each of the selected candidate sequences; providing said generated at least one amino acid sequence; Including, Each of the generated amino acid sequences is suitable for the production of a respective vaccine, each vaccine comprising at least one of the group consisting of: i) a protein defined by the generated amino acid sequence, ii) a nucleic acid capable of producing a protein defined by the generated amino acid sequence, and iii) a delivery vehicle capable of producing a protein defined by the generated amino acid sequence.

2. 2. The method of claim 1, wherein generating a plurality of reduced dimensionality arrays comprises creating a representation of the wild-type amino acid sequence using a variational autoencoder that predicts a mean and variance of input data.

3. The method of claim 2 , wherein each of the reduced dimensional arrays includes a respective group of values, and generating the plurality of candidate arrays in the reduced dimensional space includes sampling a distribution of values ​​of the plurality of reduced dimensional arrays.

4. The titer predictor comprises: receiving as input: i) a first array in the reduced dimensional space; and ii) a second array in the reduced dimensional space; providing as an output a potency score as the candidate score, the potency score defining a measure of biological response between the first sequence and the second sequence; The method according to any one of claims 1 to 3, configured to:

5. The method of any one of claims 1 to 4, wherein selecting the at least one candidate sequence as a selected candidate sequence comprises selecting the N candidate sequences having the highest candidate scores.

6. 6. The method of claim 5, wherein N has a value of 1 such that a single candidate sequence is selected.

7. The method of claim 5, wherein N is greater than 1 such that multiple candidate sequences are selected.

8. 8. The method of claim 1, wherein selecting at least one candidate sequence as a selected candidate sequence comprises selecting candidate sequences whose respective candidate scores are greater than a threshold.

9. 9. The method of any one of claims 1 to 8, wherein each of the generated amino acid sequences is different from any of the wild-type amino acid sequences.

10. The method of any one of claims 1 to 9, wherein at least one of the candidate sequences is in the plurality of reduced dimensionality sequences.

11. The method according to any one of claims 1 to 10, wherein each of said vaccines is against one of the group consisting of i) influenza, ii) human rhinovirus, iii) HIV, and iv) coronavirus diseases.

12. 1. A system for generating an amino acid sequence, comprising: one or more processors; a computer memory storing instructions that, when executed by the processor, cause the processor to perform operations; and the operation includes: receiving one or more data objects defining a plurality of wild type amino acid sequences; generating a plurality of reduced-dimensional arrays in a reduced-dimensional space from the one or more data objects, each reduced dimension array comprises data for at least one respective one of said wild-type amino acid sequences; the reduced dimensional space is less dimensional than the wild-type amino acid sequence; the plurality of reduced dimensional arrays defining a distribution of values ​​along each dimension of the reduced dimensional space; Generating, generating a plurality of candidate sequences in the reduced dimensional space using the plurality of reduced dimensional sequences; receiving one or more data objects defining a viral amino acid sequence; generating at least one reduced-dimensionality viral sequence within the reduced-dimensionality space; providing each of said candidate sequences and at least one of said reduced dimensionality viral sequences as inputs to a titer predictor; receiving a candidate score for each of the candidate sequences as output from a potency predictor; selecting at least one candidate sequence from among the candidate sequences; generating at least one new amino acid sequence for each of the selected candidate sequences; providing said generated at least one amino acid sequence; Including, Each of the generated amino acid sequences is suitable for the production of a respective vaccine, each vaccine comprising at least one of the group consisting of: i) a protein defined by the generated amino acid sequence, ii) a nucleic acid capable of producing the protein defined by the generated amino acid sequence, and iii) a delivery vehicle capable of producing the protein defined by the generated amino acid sequence.

13. 13. The system of claim 12, wherein generating the plurality of reduced dimensionality arrays comprises creating a representation of the wild-type amino acid sequence using a variational autoencoder that predicts a mean and variance of the input data.

14. 14. The system of claim 12 or 13, wherein each of the reduced dimensional arrays includes a respective group of values, and generating the plurality of candidate arrays in the reduced dimensional space includes sampling a distribution of values ​​of the plurality of reduced dimensional arrays.

15. The titer predictor comprises: receiving as input: i) a first array in the reduced dimensional space; and ii) a second array in the reduced dimensional space; providing as an output a potency score as the candidate score, the potency score defining a measure of biological response between the first sequence and the second sequence; The system according to any one of claims 12 to 14, configured to:

16. The system of any one of claims 12 to 15, wherein selecting the at least one candidate sequence as a selected candidate sequence comprises selecting the N candidate sequences having the highest candidate scores.

17. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform operations, the operations including: receiving one or more data objects defining a plurality of wild type amino acid sequences; generating a plurality of reduced-dimensional arrays in a reduced-dimensional space from the one or more data objects, each reduced dimension array comprises data for at least one respective one of said wild-type amino acid sequences; the reduced dimensional space is less dimensional than the wild-type amino acid sequence; the plurality of reduced dimensional arrays defining a distribution of values ​​along each dimension of the reduced dimensional space; Generating and generating a plurality of candidate sequences in the reduced dimensional space using the plurality of reduced dimensional sequences; receiving one or more data objects defining a viral amino acid sequence; generating at least one reduced-dimensionality viral sequence within the reduced-dimensionality space; providing each of said candidate sequences and at least one of said reduced dimensionality viral sequences as inputs to a titer predictor; receiving a candidate score for each of the candidate sequences as output from a potency predictor; selecting at least one candidate sequence from among the candidate sequences; generating at least one new amino acid sequence for each of the selected candidate sequences; providing said generated at least one amino acid sequence; Including, Each of the generated amino acid sequences is suitable for the production of a respective vaccine, each vaccine comprising at least one of the group consisting of: i) a protein defined by the generated amino acid sequence, ii) a nucleic acid capable of producing the protein defined by the generated amino acid sequence, and iii) a delivery vehicle capable of producing the protein defined by the generated amino acid sequence.

18. 20. The medium of claim 17, wherein generating a plurality of reduced dimensionality arrays comprises creating a representation of the wild-type amino acid sequence using a variational autoencoder that predicts a mean and variance of input data.

19. 20. The medium of claim 17 or 18, wherein each of the reduced dimensional arrays includes a respective group of values, and generating the plurality of candidate arrays in the reduced dimensional space includes sampling a distribution of values ​​of the plurality of reduced dimensional arrays.

20. The titer predictor comprises: receiving as input: i) a first array in the reduced dimensional space; and ii) a second array in the reduced dimensional space; providing as an output a potency score as the candidate score, the potency score defining a measure of biological response between the first sequence and the second sequence; The medium according to any one of claims 17 to 19, configured to perform the following:

Images