Methods and systems for t cell receptor (TCR) assay design

Abstract

A system and method of designing T cell receptor (TCR) assays is disclosed, including processor-based predictive modeling using HLA binding classifiers, T cell responses, sequencing T cells, and TCR classifier / regression. In particular, embodiments include feeding representations of various peptides into a trained HLA binding classifier model configured to determine an average binding prediction of overlapping peptides at each location of a virus or oncoprotein. Based on the average binding prediction, one or more peptide pools may be selected and fed into a T-cell response model along with representative blood samples associated with the patient / population of patients. In addition, the resulting sequenced T-cell responses can be used to detect T-cell response patterns. These detected patterns may be used to train a TCR classifier / regression model to predict or estimate patient state. Finally, a minimum set of detected T-cell receptors can be used to design primers for classifying or estimating patient status.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to U.S. Application Serial No. 18 / 601,946, filed March 11, 2024, entitled “Method and System for T-Cell Receptor (TCR) Assay Design,” and U.S. Provisional Application Serial No. 63 / 489,413, filed March 9, 2023, entitled “Method and System for T-Cell Receptor (TCR) Assay Design.” This application also relates to co-owned U.S. Patent Application Serial No. 17,670,385, filed February 11, 2022, entitled “HLA Clusters, Global Frequencies, and Binding Across SARS-CoV-2 Variation,” which is currently co-pending. These applications are incorporated herein by reference in their entirety. Technical Field

[0003] This disclosure generally relates to T-cell receptor (TCR) assays, and more specifically to the use of computer-based predictions to determine TCR assays.

[0004] References to sequence lists

[0005] This application contains an electronic sequence list. The sequence list file, titled N1077-10078WO01_ST26.xml, was created on March 8, 2023, and is 1,496 bytes in size. The information in the electronic sequence list is incorporated herein by reference in its entirety. Background Technology

[0006] The human immune system comprises a network of biological processes that protect humans from bacteria, microorganisms, viruses, toxins, parasites, and diseases. The immune system detects and responds to a wide variety of pathogens, from viruses to cancer cells, thus distinguishing foreign substances from healthy tissue.

[0007] Viruses contain segments of DNA or RNA enveloped in a protective protein coat. When a virus or bacteria invades the human body, it can replicate itself to cause infection or disease. When a virus encounters a human cell, it can infect the cell by attaching itself to the cell wall and injecting its viral DNA into the cell. The viral DNA can cause the cell to multiply into new viral particles. In some cases, the viral DNA causes the infected cell to eventually die and rupture, releasing new viral particles. In other cases, the infected cell may remain alive, but the viral DNA may cause viral particles to germinate from the cell.

[0008] The immune system uses white blood cells to identify and destroy infected cells. The major histocompatibility complex (MHC) (also known as human leukocyte antigen (HLA)) allows white blood cells to distinguish between healthy, natural cells and cells infected by external viruses or bacteria. MHC protein molecules label specific white blood cells (T lymphocytes, or "T cells") to detect viral infection. Specifically, MHC protein molecules present fragments of proteins (peptides) belonging to invading viruses on the cell surface to highlight the presence of infection. When a T cell recognizes a peptide on the surface of an infected cell, it can bind to the cell and destroy it or attempt to heal it. In contrast, T cells typically do not respond to healthy cells because the MHC protein molecules on healthy cells present peptides belonging to the cell itself (called self-peptides).

[0009] MHC protein molecules are mainly divided into two classes: class I and class II, which are distributed on the membranes of somatic cells. In humans, these MHC protein molecules are encoded by several genes clustered in a region on chromosome 6. The HLA corresponding to MHC class I (referred to as "HLA-I" in this article) presents intracellular peptides. For example, when a cell is infected by a virus, the HLA system carries viral fragments to the cell surface so that the immune system can destroy the cell. The HLA corresponding to MHC class II (referred to as "HLA-II" in this article) presents antigens of extracellular proteins from outside the cell to T lymphocytes. These antigens stimulate the proliferation of T helper cells (also known as CD4+ T cells). CD4+ T cells play a major role in initiating and shaping adaptive immune responses, for example by stimulating antibody-producing B cells to produce antibodies against that specific antigen. Epitopes are parts of an antigen that can bind to antibodies and be recognized by the immune system.

[0010] Antibodies are Y-shaped proteins produced by white blood cells to help eliminate viruses or help avoid the effects of viral or bacterial infections. The ends of the forked Y-shaped branches of these proteins can respond to and bind to specific antigens (e.g., bacteria, viruses, or toxins). When an antibody binds to the outer layer of a viral particle or the cell wall of a bacterium, it can prevent the virus or bacteria from moving through the human cell wall. Alternatively, large numbers of antibodies can bind to antigens and signal the complement system (a series of proteins produced in the liver) that the invader needs to be removed.

[0011] Vaccination helps the body produce its own antibodies to fight infection. While many vaccines exist that can cure diseases, coronaviruses and influenza are two examples of viral and bacterial infections that cannot currently be completely cured by vaccines. These types of viruses tend to mutate rapidly and / or have too many different strains to provide complete protection in all cases. In some situations, vaccines for coronaviruses and influenza may be a good way to avoid the effects of specific viral strains.

[0012] In light of the latest global pandemic, medical researchers are focusing on rapidly characterizing severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus that caused the global COVID-19 pandemic, in order to identify potential target proteins or peptides in order to produce vaccines that can provide therapeutic treatments.

[0013] SARS-CoV-2 has a single-stranded, positive-sense RNA genome of approximately 30 kilobases (kb), containing open reading frames encoding non-structural replicase polyproteins and structural proteins (i.e., spike protein (S), envelope protein (E), membrane protein (M), and nucleocapsid protein (N)). This positive-sense genome can serve as messenger RNA, which can be directly translated into viral proteins by the host cell's ribosomes.

[0014] Early results from research throughout 2020 indicated that the SARS-CoV-2 spike protein (S) and nucleocapsid protein (N) exhibited the highest HLA-I / -II binding recognition. Some researchers observed that the SARS-CoV-2 S and N proteins possessed the most promising candidate T and B cell epitopes. This study used reference “Wuhan-Hu-1” viral strain proteins and was based on conserved epitopes from SARS-CoV (the 2003 SARS virus) and SARS-CoV-2 predictions across 12 HLA-I alleles (determined using NetMHC4.0pan). T cell epitopes with high sequence identity to SARS-CoV-2 were independently identified using two methods.

[0015] Other researchers have observed that genetic variability across three MHC class I genes (HLA A, B, and C) may influence susceptibility and severity of SARS-CoV-2. They performed computer simulations to analyze the viral peptide-MHC class I binding affinity across 145 HLA-A, -B, and -C genotypes for all SARS-CoV-2 peptides and explored the potential for cross-protective immunity conferred by previous exposure to four common human coronaviruses. The analysis revealed 48 highly conserved amino acid sequences across 34 different coronaviruses (ORF1ab, S, E, M, and N proteins), while 56 HLAs showed no affinity for the conserved peptides. The analysis also indicated that the SARS-CoV-2 proteome was successfully sampled and presented by multiple HLA alleles. However, HLA-B*46:01 showed the fewest predicted binding peptides for SARS-CoV-2, suggesting that individuals carrying this allele may be particularly susceptible to COVID-19, as previous studies have shown their susceptibility to SARS-CoV. Conversely, HLA-A*02:02, HLA-B*15:03, and HLA-C*12:03 exhibited the strongest ability to present highly conserved SARS-CoV-2 peptides shared in common human coronaviruses, suggesting they can confer T-cell-based cross-protective immunity. The global distribution of HLA types was also reported, and the potential epidemiological impacts in the context of the COVID-19 pandemic were discussed.

[0016] Another strategy used by the researchers was to use a "megapool" of peptides predicted by HLA-I and II to identify circulating SARS-CoV-2-specific CD8 in approximately 70% and 100% of COVID-19 convalescent patients, respectively. + and CD4 + T cells. CD4 + T cell responses to the S protein (the primary target of most vaccine efforts) are robust and correlate with the levels of anti-SARS-CoV-2 IgG and IgA. The M, S, and N proteins each account for a significant portion of total CD4+. + The response rate is 11%-27%, with the remaining responses typically targeting nsp3, nsp4, ORF3a, and ORF8, etc. For CD8... + T cells recognize S and M proteins and target at least eight SARS-CoV-2 ORFs. Furthermore, SARS-CoV-2 reactive CD4+ cells were detected in approximately 40%–60% of unexposed individuals. + T cells, indicating cross-reactive T cell recognition between circulating "common cold" coronaviruses and SARS-CoV-2.

[0017] One proposed SARS-CoV-2 vaccine design concept is based on the identification of highly conserved regions of the viral genome and newly acquired adaptations, both of which predict the generation of epitopes presented on MHC classes I and II in the vast majority of the human population. Utilizing this concept, genomic regions that produce peptides highly different from the human proteome are prioritized. These are also predicted to generate B-cell epitopes. Researchers have proposed 65 33-mer peptide sequences, predicted to induce long-term immunity in most individuals, some of which can be tested via DNA or mRNA delivery strategies. These peptides include those contained within the evolutionary divergence region of the spike (S) protein—a region reportedly enhancing infectivity by strengthening binding to the ACE2 receptor—and also peptides located within the newly evolved furin cleavage site—a site thought to enhance membrane fusion.

[0018] Against the backdrop of these efforts, artificial neural networks (ANNs) (such as recurrent neural networks (RNNs)) have been successfully used in recent years for many tasks involving sequence data, where RNNs must find associations between long input sequences and output sequences, such as for predicting the binding between intact peptides and HLA protein sequences. Attention mechanisms, which enhance performance on many tasks, are an indispensable part of modern RNN networks. Attention mechanisms allow RNNs to focus on specific parts of the input sequence when predicting specific parts of the output sequence, thus making the learning process easier and the prediction quality higher.

[0019] However, to date, existing technologies have provided limited information on how the binding of HLA-I / II to SARS-CoV-2 proteins varies across viral strains and populations. In particular, current technologies have not fully revealed the association between HLA-I / II clusters, their global frequency, and binding in SARS-CoV-2 variants. For example, vaccine researchers developing SARS-CoV-2 vaccines or therapeutics have yet to find effective techniques to minimize the possibility of missing uniquely functional HLA clusters. Without technologies that provide such information, medical researchers will find it difficult to validate and implement vaccine or therapeutic concepts specifically targeting SARS-CoV-2 vulnerabilities and eliciting a strong adaptive immune response in the vast majority of the global population. Existing technologies also offer limited options for accurately tracking patient recovery or predicting the progression of SARS-CoV-2 or other viral infections. Summary of the Invention

[0020] In response to the challenges mentioned above, this article describes systems, methods, and articles for designing T-cell receptor (TCR) assays for classifying or estimating patient status.

[0021] In embodiments, a system and method are provided for designing TCR assays to classify and / or estimate patient status. One system for designing TCR assays includes processor-based predictive modeling using an HLA binding classifier, T-cell responses, sequenced T cells, and a TCR classifier / regression. Specifically, for some embodiments, the method may include training an artificial neural network (ANN), such as a convolutional neural network (CNN) or a recurrent neural network (RNN), that defines a pan-human leukocyte antigen (HLA) binding classifier model to determine the average binding prediction of overlapping peptides at each location of a viral or oncoprotein. Multiple inputs representing various peptides can be fed into the trained HLA binding classifier model. Based on the average binding prediction, one or more peptide pools can be selected. Furthermore, one or more peptide pools and multiple inputs associated with multiple blood samples associated with a patient or patient population can be fed into a T-cell response model. The resulting T-cell responses can be sequenced using a sequencer. One or more T-cell response patterns can be detected from the sequenced T-cell responses. The TCR classifier / regression model can be trained to predict or estimate patient status based on the detected one or more T-cell response patterns. In some embodiments, a minimal set of T-cell receptors can be detected. Ultimately, primers (defined by TCR assays) can be designed using the detected minimal set of T-cell receptors for classifying or estimating patient status.

[0022] In some embodiments, a system for designing TCR assays is provided. A cloud-based TCR assay system may include a processor coupled to memory, a storage unit, and a processor-based TCR assay module (coupled with an ANN model generator to generate an HLA binding classifier model, a T-cell response model, and a TCR classifier / regression model). The HLA binding classifier model is configured to determine the average binding prediction of overlapping peptides at each location of a viral or oncoprotein. The TCR assay module may further include a peptide unit coupled to the HLA binding classifier model to feed multiple inputs representing various peptides into the trained HLA binding classifier model. Using the peptide unit, one or more peptide pools can be selected based on the average binding prediction. A sequencer may be included within the TCR assay module coupled to the T-cell response model. The sequencer is designed to supply multiple inputs associated with multiple blood samples associated with a patient or patient population, which can be fed into the T-cell response model. The sequencer is also configured to sequence T-cell receptor responses. One or more T-cell response patterns can be detected from the sequenced T-cell responses. The TCR classifier / regression model can be configured to detect one or more T-cell response patterns. Furthermore, the TCR classifier / regression model can be trained to predict or estimate patient status based on one or more detected T cell response patterns. In some embodiments, the TCR classifier / regression model can detect a minimal set of T cell receptors for classifying or estimating patient status. The TCR assay module may further include a primer agent to design primers using the minimal set of detected T cell receptors for classifying or estimating patient status.

[0023] In some embodiments, instructions are provided on a tangible, non-transitory computer-readable medium that, when executed by a processor, cause the processor to perform the TCR assay design method described herein. In some embodiments, a method for designing a TCR assay is provided. Specifically, some embodiments may include training an ANN (e.g., a CNN or RNN) to define an HLA binding classifier model to determine the average binding prediction of overlapping peptides at each location of a viral or oncoprotein. Multiple inputs representing various peptides may be fed into the trained HLA binding classifier model. Based on the average binding prediction, one or more peptide pools may be selected. Furthermore, one or more peptide pools and multiple inputs associated with multiple blood samples associated with a patient or patient population may be fed into a T-cell response model. The resulting T-cell responses may be sequenced using a sequencer. One or more T-cell response patterns may be detected from the sequenced T-cell responses. The TCR classifier / regression model may be trained to predict or estimate patient status based on the detected one or more T-cell response patterns. In some embodiments, a minimal set of T-cell receptors may be detected. Finally, primers (defining the TCR assay) may be designed using the detected minimal set of T-cell receptors for classifying or estimating patient status.

[0024] In some embodiments, the viral or oncoprotein is encoded as a variable-length peptide. The cancer or viral protein may comprise a SARS-CoV-2 protein variant. A SARS-CoV-2 protein variant may comprise a SARS-CoV-2 nucleocapsid (N) protein variant. In other instances, the SARS-CoV-2 protein variant comprises a SARS-CoV-2 spike (S) protein variant.

[0025] In some embodiments, determining the average binding prediction includes classifying a peptide as a binder when the average binding prediction corresponding to the peptide meets a binding value threshold. The TCR assay design method may further include selecting one or more peptide pools to focus on one or more of the following: a specific site, hotspot, or receptor-binding domain of the virus or oncoprotein. In other embodiments, one or more peptide pools may be selected to focus on multiple regions or hotspots of the virus or oncoprotein. One or more peptide pools may also be selected to focus on the entire virus or oncoprotein. Furthermore, one or more peptide pools may be selected based on at least one of CD4 or CD8 T cell interactions. In some embodiments, one or more peptide pools may be selected based at least on the average binding prediction of the HLA-I functional group. Furthermore, one or more peptide pools may be selected based at least on the average binding prediction of the HLA-II functional group. One or more peptide pools may also be selected based on regions of predicted binding frequency in the HLA-I and HLA-II functional groups. One or more peptide pools may be selected based on pan-HLA binding prediction.

[0026] In some embodiments, the T-cell response assay includes at least one of the following: enzyme-linked immunosorbent assay (ELISpot) assay, cytotoxic T lymphocyte (CTL) assay, and DNA barcoded peptide-MHC (pMHC) multimer assay. Additionally, the T-cell response assay may include a T-cell response assay based on a synthetic TCR.

[0027] In some embodiments, the synthetic TCR assay is intended to supplement T-cell response data for the patient or patient population. Furthermore, the TCR assay can be used to classify or estimate patient status. In some instances, patient status includes determining whether the patient has a medical condition. Patient status may also include an estimate of the patient's medical outcomes. Additionally, patient status may include an estimate of the patient's disease progression. In some embodiments, therapeutic treatment may be administered to the patient based on the classified or estimated patient status.

[0028] Other aspects and advantages of the embodiments will become apparent from the following detailed description taken in conjunction with the accompanying drawings, which illustrate the principles of the described embodiments by way of example. Attached Figure Description

[0029] The described embodiments and their advantages can be best understood by referring to the following description taken in conjunction with the accompanying drawings. These drawings are in no way intended to limit any changes in form and detail that may be made to the described embodiments by those skilled in the art without departing from the spirit and scope of the described embodiments.

[0030] Figure 1 This is a block diagram of an exemplary network incorporating systems and methods for designing TCR measurements, according to some embodiments.

[0031] Figure 2 According to some embodiments, it is used for Figure 1 A block diagram of an exemplary system for TCR measurement within an exemplary network component.

[0032] Figure 3 According to some embodiments Figure 1 A block diagram of an exemplary TCR determination agent within an exemplary network component.

[0033] Figure 4 This is an exemplary flowchart of a method for designing a TCR measurement according to some embodiments.

[0034] Figure 5 This is an illustration of an exemplary computing device that can implement the embodiments described herein. Detailed Implementation

[0035] The following examples describe systems and methods for designing T-cell receptor (TCR) assays. Those skilled in the art will understand that the examples can be practiced without some or all of these specific details. In other instances, well-known procedures have not been described in detail so as not to unnecessarily obscure the examples.

[0036] In some embodiments, systems and methods for designing T-cell receptor (TCR) assays include processor-based predictive modeling using an HLA binding classifier, T-cell responses, sequencing T cells, and a TCR classifier / regressor. Specifically, some embodiments may include training an artificial neural network, such as a convolutional neural network (CNN) or a recurrent neural network (RNN), to define a pan-human leukocyte antigen (HLA) binding classifier model to determine the average binding prediction of overlapping peptides at each location of a viral or oncoprotein. Multiple inputs representing various peptides can be fed into the trained HLA binding classifier model. One or more peptide pools are selected based on the average binding prediction. Furthermore, one or more peptide pools and multiple inputs associated with multiple blood samples associated with a patient or patient population can be fed into a T-cell response model. The resulting T-cell responses can be sequenced using a sequencer. One or more T-cell response patterns can be detected from the sequenced T-cell responses. A TCR classifier / regression model can be trained to predict or estimate patient status based on the detected one or more T-cell response patterns, and primers can be designed using a minimal set of detected T-cell receptors for classifying or estimating patient status.

[0037] In some embodiments, the viral or oncoprotein is encoded as a variable-length peptide. The cancer or viral protein may comprise a SARS-CoV-2 protein variant. A SARS-CoV-2 protein variant may comprise a SARS-CoV-2 nucleocapsid (N) protein variant. In other instances, the SARS-CoV-2 protein variant comprises a SARS-CoV-2 spike (S) protein variant.

[0038] In some embodiments, determining the average binding prediction includes classifying a peptide as a binder when the average binding prediction corresponding to the peptide meets a binding value threshold. The TCR assay design method may further include selecting one or more peptide pools to focus on one or more of the following: a specific site, hotspot, or receptor-binding domain of the virus or oncoprotein. In other embodiments, one or more peptide pools may be selected to focus on multiple regions or hotspots of the virus or oncoprotein. One or more peptide pools may also be selected to focus on the entire virus or oncoprotein. Furthermore, one or more peptide pools may be selected based on at least one of CD4 or CD8 T cell interactions. In some embodiments, one or more peptide pools may be selected based at least on the average binding prediction of the HLA-I functional group. Furthermore, one or more peptide pools may be selected based at least on the average binding prediction of the HLA-II functional group. One or more peptide pools may also be selected based on regions of predicted binding frequency in the HLA-I and HLA-II functional groups. One or more peptide pools may be selected based on pan-HLA binding prediction.

[0039] In some embodiments, the T-cell response assay includes at least one of the following: enzyme-linked immunosorbent assay (ELISpot) assay, cytotoxic T lymphocyte (CTL) assay, and DNA barcoded peptide-MHC (pMHC) multimer assay. Additionally, the T-cell response assay may include a T-cell response assay based on a synthetic TCR.

[0040] In some embodiments, the synthetic TCR assay is intended to supplement T-cell response data for the patient or patient population. Furthermore, the TCR assay can be used to classify or estimate patient status. In some instances, patient status includes determining whether the patient has a medical condition. Patient status may also include an estimate of the patient's medical outcomes. Additionally, patient status may include an estimate of the patient's disease progression. In some embodiments, therapeutic treatment may be administered to the patient based on the classified or estimated patient status.

[0041] Advantageously, the systems and methods designed for TCR assays enable the tracking of the progression of viral infection in patients. In particular, methods designed for TCR assays can detect the progression of infection based on T-cell responses, given blood sample data associated with patients or patient populations.

[0042] Numerous details are set forth in the following description. However, it will be apparent to those skilled in the art that the invention can be practiced without these specific details. In some instances, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring the invention.

[0043] Some parts of the following description are presented based on algorithms and symbolic representations of operations on data bits within computer memory. These algorithmic descriptions and representations are means by which those skilled in the art of data processing most effectively communicate the essence of their work to others skilled in the art. An algorithm is hereby and generally conceived as a series of self-consistent steps that produce a desired result. These steps require physical manipulation of physical quantities. Typically (but not necessarily), these physical quantities exist in the form of electrical or magnetic signals, capable of being stored, transmitted, combined, compared, and otherwise manipulated. For the main reasons of general expression, these signals are sometimes referred to as bits, values, elements, symbols, characters, terms, numbers, etc., for convenience.

[0044] However, it should be remembered that all these and similar terms will be associated with appropriate physical quantities and are merely convenient labels applied to those quantities. Unless otherwise explicitly stated, it will be apparent from the following discussion that it should be understood that throughout this specification, when terms such as “provide,” “generate,” “install,” “monitor,” “execute,” “receive,” “record,” and “intercept” are used, the discussion refers to the actions and processes of a computer system or similar electronic computing device that manipulates and transforms data represented as physical (electronic) quantities in computer system registers and memories into other data also represented as physical quantities in computer system memory or registers or other such information storage, transmission, or display devices.

[0045] Various embodiments also relate to an apparatus for performing the operations described herein. This apparatus may be specifically constructed for particular needs or may include a general-purpose computer selectively activated or reconfigured by a computer program stored in the computer. Such a computer program may be stored in a computer-readable storage medium, such as, but not limited to, any type of disk (including floppy disks, optical disks, CD-ROMs, and magneto-optical disks), read-only memory (ROM), random access memory (RAM), EPROM, EEPROM, magnetic cards or optical cards, or any type of medium suitable for storing electronic instructions, each coupled to a computer system bus.

[0046] The reference to "one embodiment" or "an embodiment" in this specification means that a particular feature, structure, or characteristic described in connection with that embodiment is included in at least one embodiment of the invention. The phrase "in one embodiment" in various places throughout this specification does not necessarily refer to the same embodiment. Throughout the description of the drawings, the same reference numerals denote the same elements.

[0047] The various techniques described herein are improvements upon current technologies to provide insights into the associations between HLA-I / II clusters, global frequencies, and binding regions in SARS-CoV-2 variants. In particular, these techniques facilitate the identification of missing, functionally unique HLA clusters in the development of vaccines or antiviral therapeutics. These techniques also provide precise tracking of patient recovery progress and / or prediction of viral infection progression. It should be understood that the various embodiments can be implemented in multiple ways, for example, by process, apparatus, system, device, method, or by combinations thereof. Several embodiments of the invention are described below.

[0048] refer to Figure 1 An exemplary network is shown that incorporates systems and methods for designing T-cell receptor (TCR) assays. As shown, the exemplary network architecture 100 may include at least one client node (computing device) 110, 112, and 114 communicating with a server 150 via a network 140. As described above, all or part of the network architecture 200 may perform, individually or in combination with other elements, one or more steps disclosed herein (e.g., Figure 4 (One or more of the steps shown in the diagram). All or part of the network architecture 100 may also be used to perform and / or as a means to perform other steps and features set forth in this disclosure. In one instance, computing device 110 may be programmed to have one or more of agents 300 (described in detail below). Additionally or alternatively, server 150 may be programmed to have one or more of modules 200. Although not shown, in various embodiments, client nodes (110, 112, and 114) including TCR determination agents 300 may be notebook computers, desktop computers, microprocessor-based or programmable consumer electronics, network devices, mobile phones, smartphones, pagers, radio frequency (RF) devices, infrared (IR) devices, personal digital assistants (PDAs), set-top boxes, cameras, integrated devices combining at least two of the foregoing devices, etc.

[0049] In some embodiments, a TCR assay agent 300 having a peptide unit 340, a sequencer 350, and a primer agent 360 can act as an apparatus for communicating with server 150 to perform a method for designing a real-time TCR assay, as described in more detail below. In other embodiments, a TCR assay module 200 having a TCR assay design process utilizing predictive modeling can communicate with each client node 110, 112, and 114 and act as the sole agent for performing the method for designing a TCR assay as described herein. Client nodes 110, 112, and 114, server 150, and storage device 160 can reside on the same LAN or on different LANs that can be coupled together via the Internet but are separated by firewalls, routers, and / or other network devices. In one embodiment, client nodes 110, 112, and 114 can be coupled to network 140 via a mobile communication network. In another embodiment, client nodes 110, 112, and 114, server 150, and storage device 160 can reside on different networks. In some embodiments, server 150 can reside in a cloud network. Although not shown, in various embodiments, client nodes 110, 112, and 114 may be notebook computers, desktop computers, microprocessor-based or programmable consumer electronics, network devices, mobile phones, smartphones, pagers, radio frequency (RF) devices, infrared (IR) devices, personal digital assistants (PDAs), set-top boxes, cameras, integrated devices combining at least two of the aforementioned devices, etc. In some embodiments, each client node may include a TCR measurement module 230, which may operate fully or partially to perform a TCR measurement design (client nodes 110, 112, and 114) according to the methods disclosed herein.

[0050] TCR assay server 150 may include a processor (not shown), memory (not shown), and TCR assay system 200 (with TCR assay module 230). In some embodiments, server 150 may include processing software instructions and / or hardware logic required for TCR assay design according to embodiments described herein. Server 150 may provide remote cloud storage capabilities for call classification, call filtering, and various types of associated security policies via storage device 160 coupled via network 140. Furthermore, server 150 may provide remote storage capabilities for storing AI model data, peptide data, T-cell response data, and blood sample data. Moreover, server 150 may be coupled to one or more fabrication devices (not shown) or any other secondary data storage device. Thus, databases of patient profile data and user policy data may be stored in local data storage, remote disks, secondary data storage devices, or fabrication devices (not shown). In some embodiments, client nodes 110, 112, and 114 may retrieve past results related to peptide pools, T-cell responses, and blood sample data from remote data storage devices to local data storage device 158. In other embodiments, the database of AI strategies, previous TCR measurement results, etc., may be stored locally on one or more of client nodes 110, 112, and 114 or server 150. For remote storage purposes, local data storage unit 160 may be one or more centralized data repositories with a mapping between each piece of data and its corresponding location within a remote storage device. The local data storage device may represent a single or multiple data structures (databases, repositories, files, etc.) residing on one or more mass storage devices (such as magnetic or optical storage-based disks, tapes, or hard disks). The local data storage device may be an internal component of server 150. Alternatively, local data storage device 160 may also be as follows: Figure 1 The device is externally coupled to server 150 or remotely coupled via a network. Furthermore, server 150 can communicate with remote storage devices via public or private networks. Although not shown, in various embodiments, server 150 may be a laptop computer, desktop computer, microprocessor-based or programmable consumer electronics, network equipment, mobile phone, smartphone, radio frequency (RF) device, infrared (IR) device, personal digital assistant (PDA), set-top box, integrated device combining at least two of the aforementioned devices, etc.

[0051] Client nodes 110, 112, and 114 typically represent any type or form of computing device or system, such as Figure 5The exemplary computing system 500 is shown in the example. Similarly, server 150 generally refers to a computing device or system configured to provide various database services and / or run certain software applications, such as an application server or database server. Network 140 generally refers to any telecommunications or computer network, including, for example, an intranet, WAN, LAN, PAN, or the Internet. For embodiments, client nodes 110, 112, and 114 and / or server 150 may include Figure 2 All or part of the Chinese system 200.

[0052] In some embodiments, one or more storage devices (not shown) may be directly attached to server 150. A storage device generally refers to any type or form of storage device or medium capable of storing data and / or other computer-readable instructions. In some embodiments, the storage device may refer to a network access storage (NAS) device configured to communicate with server 150 using various protocols such as Network File System (NFS), Server Message Block (SMB), or Common Internet File System (CIFS).

[0053] Server 150 can also be connected to a Storage Area Network (SAN) structure (not shown). A SAN structure generally refers to any type or form of computer network or architecture that facilitates communication between multiple storage devices. The SAN structure can facilitate communication between server 150 and multiple storage devices (not shown) and / or intelligent storage arrays (not shown). The SAN structure can also facilitate communication between client nodes 110, 112, and 114 and storage devices and / or intelligent storage arrays via network 140 and server 150 in such a way that devices 170(1)-(N) and array 180 appear as local access devices for client nodes 110, 112, and 114.

[0054] In some embodiments, and with reference to Figure 5 An exemplary computing system 500 includes a communication interface for providing connectivity between each client node 110, 112, and 114 and network 150. Client nodes 110, 112, and 114 are configured to access information from a database coupled to server 150 using, for example, a web browser or other client software. Such software may allow client nodes 110, 112, and 114 to access data hosted by server 150, local storage devices, remote storage devices, or intelligent storage arrays. Although... Figure 1 The use of a network (such as the Internet) to exchange data is depicted, but the embodiments described and / or shown herein are not limited to the Internet or any particular network-based environment.

[0055] In at least one embodiment, all or a portion of one or more of the exemplary embodiments disclosed herein may be encoded as a computer program and loaded into and executed by: server 150, local storage device, remote storage device, or smart storage array, or any combination thereof. All or a portion of one or more of the exemplary embodiments disclosed herein may also be encoded as a computer program, stored in server 150, and distributed via network 140 to one or more of client nodes 110, 112, and 114.

[0056] One or more components of network architecture 100 may, individually or in combination with other elements, perform and / or as a means of performing one or more steps of the exemplary method for TCR determination design. It should be understood that the components of the exemplary operating environment 100 are exemplary, and more or fewer components may exist in various configurations. It should be understood that the operating environment may be part of a distributed computing environment, a cloud computing environment, a client-server environment, etc.

[0057] refer to Figure 2 , showed Figure 1 An exemplary embodiment of the TCR measurement design system 200 within an exemplary network component is provided. The exemplary system 200 can be implemented in various ways. For example, all or part of the exemplary system 200 may represent... Figure 1 This is a portion of the exemplary system 100. As shown in the figure, the exemplary system 200 may include a memory 210, a processor 212, and a storage database 214. The system may include one or more TCR assay modules 230 for performing one or more tasks. For example, and as will be explained in more detail below, the TCR assay module 230 may include an artificial intelligence neural network (ANN) generator 232 coupled to define an HLA binding classifier model 234, a T cell response model 236, and a TCR classifier / regression model 238. The TCR assay module 230 may further include a peptide unit 240, a sequencer 242, a primer proxy 244, and a T cell pattern detection unit 246. The peptide unit 240 is configured to store and feed a plurality of encoded peptides into the trained HLA binding classifier model 234. The sequencer 242 is configured to sequence identified responsive T cells. The T cell pattern detection unit 244 can detect one or more T cell response patterns common to patients or patient populations. Primer agent 246 is configured to design one or more primers that define a TCR assay for classifying or estimating patient status.

[0058] In operation, the TCR assay module 230 can use the ANN model generator 232 within the TCR assay module 230 to train an ANN defining a pan-human leukocyte antigen (HLA) binding classifier model 234. Using a first plurality of inputs, the trained HLA binding classifier model 234 is configured to independently determine the average binding prediction of overlapping peptides at each location on a viral or oncoprotein for each of multiple test HLAs comprising HLA-I and HLA-II functional groups. Furthermore, the peptide unit 240 can retrieve a second plurality of inputs representing viral or oncoproteins encoded as multiple peptides from local or remote storage. The peptide unit 240 can feed these inputs into the trained HLA binding classifier model 234. The HLA binding classifier model 234 can be configured to receive multiple inputs representing multiple peptides from the peptide unit 240. The peptide unit 240 can select one or more peptide pools from the multiple peptides based on the average binding prediction derived from the HLA binding classifier model 234. In some embodiments, blood samples can be retrieved from one or more of the TCR assay agents 300 within client nodes 110, 112, or 114. An ANN model generator 232 can generate a T-cell response model 236 using one or more peptide pools and a third plurality of inputs. The T-cell response model can be trained to predict the peptides or protein fragments most likely to elicit a T-cell response based on a database of validated T-cell epitopes and peptides that fail to elicit a T-cell response. Predictions from the T-cell response model can be further refined using peptide pools derived from aggregated HLA binding predictors to improve the accuracy of the proposed epitopes. A sequencer 242 can sequence samples identified from the results of T-cell response testing. A T-cell pattern detection unit 244 can detect one or more T-cell response patterns common to a patient or patient population. The ANN model generator 232 can generate a TCR classifier / regression model 238 based on at least one or more T-cell response patterns. T-cell response patterns can be identified by training a TCR classifier or regression model to distinguish disease- or patient-state-specific TCR sequences from TCR sequences common in patients who do not represent the target condition. Alternatively, patient-specific TCR patterns can be characterized nonparametrically by identifying disease-specific clusters of TCR sequences in the sequence embedding space. A trained TCR classifier or regression model 238 can determine a minimal set of T cell receptors for classifying or estimating patient status. This selection can be achieved by choosing top-ranked TCR patterns that appear in a large number of patients based on the prediction scores of the disease-specific TCR classifier. A primer agent 246 can design primers based on the determined minimal set of T cell receptors, defining TCR assays for classifying or estimating patient status.

[0059] In some embodiments, the method for designing a T-cell receptor (TCR) assay can be implemented entirely within the TCR assay system 200 on server 150. In other embodiments, the method can be implemented using a TCR assay agent 300 on client nodes (110, 112, 114) and the TCR assay system 200 (which will be referred to in the following embodiments). Figure 3 (A more detailed description) Implementation.

[0060] In some embodiments, the viral or oncoprotein is encoded as a variable-length peptide. The cancer or viral protein may comprise a SARS-CoV-2 protein variant. A SARS-CoV-2 protein variant may comprise a SARS-CoV-2 nucleocapsid (N) protein variant. In other instances, the SARS-CoV-2 protein variant comprises a SARS-CoV-2 spike (S) protein variant.

[0061] In some embodiments, determining the average binding prediction includes classifying a peptide as a binder when the average binding prediction corresponding to the peptide meets a binding value threshold. The TCR assay design method may further include selecting one or more peptide pools to focus on one or more of the following: a specific site, hotspot, or receptor-binding domain of the virus or oncoprotein. In other embodiments, one or more peptide pools may be selected to focus on multiple regions or hotspots of the virus or oncoprotein. One or more peptide pools may also be selected to focus on the entire virus or oncoprotein. Furthermore, one or more peptide pools may be selected based on at least one of CD4 or CD8 T cell interactions. In some embodiments, one or more peptide pools may be selected based at least on the average binding prediction of the HLA-I functional group. Furthermore, one or more peptide pools may be selected based at least on the average binding prediction of the HLA-II functional group. One or more peptide pools may also be selected based on regions of predicted binding frequency in the HLA-I and HLA-II functional groups. One or more peptide pools may be selected based on pan-HLA binding prediction.

[0062] In some embodiments, the T-cell response assay includes at least one of the following: enzyme-linked immunosorbent assay (ELISpot) assay, cytotoxic T lymphocyte (CTL) assay, and DNA barcoded peptide-MHC (pMHC) multimer assay. Additionally, the T-cell response assay may include a T-cell response assay based on a synthetic TCR.

[0063] In some embodiments, the synthetic TCR assay is intended to supplement T-cell response data for the patient or patient population. Furthermore, the TCR assay can be used to classify or estimate patient status. In some instances, patient status includes determining whether the patient has a medical condition. Patient status may also include an estimate of the patient's medical outcomes. Additionally, patient status may include an estimate of the patient's disease progression. In some embodiments, therapeutic treatment may be administered to the patient based on the classified or estimated patient status.

[0064] It should be understood that the components of the exemplary operating environment 100 are exemplary, and more or fewer components can exist in various configurations. It should be understood that the operating environment can be part of a distributed computing environment, a cloud computing environment, a client-server environment, etc.

[0065] As used herein, the term "module" can describe a given functional unit that can perform according to one or more embodiments of the invention. As used herein, a module can be implemented using any form of hardware, software, or a combination thereof. For example, one or more processors, ASICs, PLAs, PALs, CPLDs, FPGAs, logic components, software routines, or other mechanisms can be implemented to form a module. In implementations, the various modules described herein can be implemented as discrete modules, or the described functions and features can be shared partially or wholly among one or more modules. In other words, as will be apparent to those skilled in the art after reading this specification, the various features and functions described herein can be implemented in any given application and can be implemented in one or more separate or shared modules in various combinations and arrangements. Although various feature or functional elements can be described separately or claimed as separate modules, those skilled in the art will understand that these features and functions can be shared among one or more common software and hardware elements, and this description should not require or imply the use of separate hardware or software components to implement such features or functions.

[0066] refer to Figure 3 , showed Figure 1 The exemplary TCR determination agent 300 is located within an exemplary network component. The exemplary agent 300 can be implemented in various ways. For example, all or part of the exemplary agent 300 may represent... Figure 1 This is a portion of the exemplary system 100. More specifically, the TCR measurement agent 300 may include one or more components of the TCR measurement module 200 for local processing of the methods described herein for designing TCR measurements. In some embodiments, such as Figure 3As shown, the exemplary agent 300 may include a memory 310, a processor 320, and a storage database 330. The agent may include one or more processing modules 340 for performing one or more tasks. For example, and as will be explained in more detail below, processing module 340 may include a peptide unit 342, a sequencer 344, a T-cell pattern detection unit 346, and a primer agent 348. Similar to peptide unit 240, peptide unit 342 is configured to store and feed multiple encoded peptides to TCR assay module 230 on server 150. Figure 1 and Figure 2 The trained HLA binding classifier model 234 is used. The sequencer 344 is configured to sequence T cells identified from the results of a T cell response test using a T cell model generated by an ANN model generator 232 on server 150. The T cell pattern detection unit 346 can detect one or more T cell response patterns common to a patient or patient population based on the T cell model. When communicating with the TCR assay module 230, the primer agent 348 is configured to design one or more primers defining the TCR assay based on a minimal set of identified T cell receptors for classifying or estimating patient status.

[0067] In operation, the TCR assay module 230, in collaboration with the TCR agent 300, can train an ANN defining a pan-human leukocyte antigen (HLA) binding classifier model 234 using an ANN model generator 232 within the TCR assay module 230. Using a first plurality of inputs from peptide units 342 and local storage 330, the trained HLA binding classifier model 234 is configured to independently determine the average binding prediction of overlapping peptides at each location on a viral or oncoprotein for each of a plurality of test HLAs comprising HLA-I and HLA-II functional groupings. Furthermore, peptide units 342 can retrieve a second plurality of inputs representing viral or oncoproteins encoded as multiple peptides from local or remote storage. Peptide units 342 can feed these inputs into the trained HLA binding classifier model 234. Peptide units 342 can select one or more peptide pools from the plurality of peptides based on the average binding predictions derived from the HLA binding classifier model 234. As described above, in some embodiments, blood samples can be retrieved from one or more of the TCR assay agents 300 within client nodes 110, 112, or 114. ANN model generator 232 can generate T-cell response models 236 using one or more peptide pools and a third plurality of inputs. Sequencing instrument 344 can sequence the T-cell response models identified from the results of T-cell response testing. T-cell pattern detection unit 346 can detect one or more T-cell response patterns common to patients or patient groups. On server 150, ANN model generator 232 can generate a TCR classifier / regression model 238 based on at least one or more T-cell response patterns. The trained TCR classifier or regression model 238 can determine a minimal set of T-cell receptors for classifying or estimating patient status. While communicating with the TCR classifier or regression model 238 on server 150, primer agents 348 on any client node can design primers based on the determined minimal set of T-cell receptors, defining TCR assays for classifying or estimating patient status.

[0068] Figure 4This is an exemplary flowchart of a method for designing a TCR assay according to some embodiments. In action 405, an ANN is trained to generate an HLA binding classifier model using a first plurality of inputs. For example, an ANN model generator 232 can train the ANN using a first plurality of inputs that define a pan-human leukocyte antigen (HLA) binding classifier model 234. The trained HLA binding classifier model 234 can be configured to independently determine the average binding prediction of overlapping peptides at each location of a viral or oncoprotein for each of a plurality of test HLAs comprising HLA-I and HLA-II functional groupings. Furthermore, a second plurality of inputs can be retrieved, wherein these inputs represent viral or oncoproteins encoded as multiple peptides in action 410. For example, peptide unit 240 can retrieve a second plurality of inputs representing viral or oncoproteins encoded as multiple peptides from a local or remote storage device. The method for designing a TCR assay may further include, in action 415, feeding the second plurality of inputs representing multiple peptides into the trained HLA binding classifier model. For example, HLA binding classifier model 234 may be coupled to peptide unit 240 to receive the plurality of inputs representing multiple peptides. Furthermore, the method for designing a TCR assay may include, in action 420, selecting one or more peptide pools from a plurality of peptides based at least on average binding predictions. For example, peptide unit 240 may select one or more peptide pools from a plurality of peptides based on average binding predictions derived from HLA binding classifier model 234. Additionally, in action 425, the method may include retrieving a third plurality of inputs associated with a plurality of blood samples, wherein these blood samples represent a patient or patient population. In some embodiments, blood samples may be retrieved from one or more of the TCR assay agents 300 within client nodes 110, 112, or 114. In action 430, the method may include instantiating a T cell response model by an ANN model generator using one or more peptide pools and a third plurality of inputs. For example, ANN model generator 232 may use one or more peptide pools and a third plurality of inputs to generate a T cell response model 236. The method for designing a TCR assay may include, in action 435, sequencing responsive T cells identified from the results of a T cell response test using a sequencer. For example, sequencer 242 may sequence responsive T cells identified from the results of a T cell response test. The method may include, in action 440, detecting one or more T-cell response patterns common to the patient or patient population, based at least on data obtained from sequencing these responsive T cells. For example, T-cell response model 236 may detect one or more T-cell response patterns common to the patient or patient population. In action 445, the method may include training a TCR classifier or regression model by an ANN model generator to predict or estimate patient status using a dataset based at least on the one or more T-cell response patterns.For example, the ANN model generator 232 can generate a TCR classifier / regression model 238 based at least on one or more T cell response patterns. Furthermore, the method for designing TCR assays can include, in action 450, using a trained TCR classifier or regression model to determine a minimal set of T cell receptors for classifying or estimating patient states. For example, a trained TCR classifier or regression model can determine a minimal set of T cell receptors for classifying or estimating patient states. In action 455, the method can include designing primers based on the determined minimal set of T cell receptors, which include TCR assays for classifying or estimating patient states. For example, a primer agent can design primers based on the determined minimal set of T cell receptors, which include TCR assays for classifying or estimating patient states.

[0069] In another embodiment, the method for designing a TCR assay disclosed herein includes a process for generating an immunotherapeutic agent comprising antigen-reactive T cells. In some aspects, the method includes identifying novel epitope antigen-reactive T cells. In some aspects, the method involves generating a population of novel epitope antigen-reactive T cells using one or more peptides containing the same amino acid sequence as the novel epitope derived from a patient. PCT patent application WO / 2022 / 086727 is hereby incorporated by reference.

[0070] Other embodiments are described below.

[0071] (1) A method for designing a T-cell receptor (TCR) assay performed by a processor-based TCR assay module, the method comprising:

[0072] Obtain the first multiple inputs representing various peptides;

[0073] The first multiple inputs are used to train an artificial neural network (ANN) defining a pan-human leukocyte antigen (HLA) binding classifier model, wherein the trained HLA binding classifier model is configured to independently determine the average binding prediction of overlapping peptides at each location of a virus or oncoprotein for each of multiple test HLAs containing HLA-I and HLA-II functional groups.

[0074] Obtain a second multiple input representing a viral or oncoprotein encoded as multiple peptides;

[0075] The second plurality of inputs are fed into the trained HLA binding classifier model, wherein the trained HLA binding classifier is configured to determine the average binding prediction of overlapping peptides among the plurality of peptides.

[0076] Based on these average binding predictions, one or more peptide pools are selected from the multiple peptides;

[0077] Obtain a third multiple input associated with multiple blood samples, where these blood samples represent patients or patient groups;

[0078] A T-cell response model is instantiated based on the one or more peptide pools and the third plurality of inputs; wherein the T-cell response model is trained to predict peptides and protein fragments that are highly likely to elicit a T-cell response based on validated T-cell epitopes and peptides that fail to elicit a T-cell response.

[0079] Responsive T cells identified based on T cell response criteria were sequenced using a sequencer.

[0080] Based on data obtained from sequencing these responsive T cells, detect one or more T cell response patterns common to the patient or patient population;

[0081] Train a TCR classifier / regression model to predict or estimate patient status using a dataset based on one or more T cell response patterns;

[0082] A minimal set of T-cell receptors for classifying or estimating the patient's condition was determined using a trained TCR classifier / regression model; and

[0083] One or more primers are designed based on the identified minimal set of T-cell receptors, which define a TCR assay for classifying or estimating the patient's condition.

[0084] (2) The method as described in (1), wherein training the HLA-based classifier model includes:

[0085] Multiple test HLAs encoding variable-length proteins were obtained, wherein the multiple test HLAs comprise HLA-I and HLA-II functional groups;

[0086] The classifier model is used to process the variable-length peptides encoding the viral protein and the variable-length proteins corresponding to the multiple test HLAs, such that, independently of the test HLAs, the classifier model is operable to determine the average binding prediction of overlapping peptides at each position of the viral protein.

[0087] Independently based on HLA testing:

[0088] The aggregated average binding predictions are mapped to locations along the test viral proteins to enable indicator peptide-HLA interactions;

[0089] Use a sliding window of fixed length to determine the location of the most recent maximum value of these combined average predictions;

[0090] The region of highest value is determined by selecting the location of the most recent maximum value within the range of the highest percentage of the predicted average.

[0091] Select peptides that are classified as conjugates overlapping with the region of highest maxima; and

[0092] Determine the pan-HLA maximum region, wherein the determination includes setting unselected locations to zero, calculating the average along the average combined with the predicted HLA axis, and selecting the pan-HLA maximum value within the highest percentage of the value based on the average.

[0093] Independently for each of the HLA-I and HLA-II functional groups:

[0094] Selected peptides classified as conjugates were filtered to identify candidate peptides overlapping with these maximum maxima regions based on a summary of these pan-HLA maxima regions; and

[0095] Incorporate one or more of these candidate peptides into mRNA-based vaccines or therapeutic treatments for patients.

[0096] (3) The method as described in any one of (1)-(2), wherein training the TCR classifier / regression model to predict or estimate the patient status comprises:

[0097] Distinguish between patient-state-specific TCR sequences and general TCR sequences associated with patients who do not represent the target disease associated with that patient state;

[0098] Based on this differentiation, T cell response patterns can be identified; and

[0099] This TCR classifier / regression model is generated based on the identified T cell response patterns.

[0100] (4) The method of any one of (1)-(3), wherein training the TCR classifier / regression model to predict or estimate the patient status comprises:

[0101] Distinguish between patient-state-specific TCR sequences and general TCR sequences associated with patients who do not represent the target disease associated with that patient state;

[0102] Based on this distinction, TCR sequences in the sequence embedding space associated with the target disease are identified; and

[0103] The TCR classifier / regression model is generated using the identified TCR sequences.

[0104] (5) The method of any one of (1)-(4), wherein determining the minimal set of T cell receptors comprises:

[0105] Retrieve predicted scores from trained TCR classifiers / regression models for multiple patients; and

[0106] Select one or more TCR modes based on the retrieved prediction scores.

[0107] (6) The method of any one of (1)-(5), wherein the method further comprises selecting the one or more peptide pools based on one or more of the following: a specific site, hotspot or receptor-binding domain of the virus or oncoprotein.

[0108] (7) The method of any one of (1)-(6), wherein the method further comprises selecting the one or more peptide pools based on multiple regions or hotspots of the virus or oncoprotein.

[0109] (8) The method of any one of (1)-(7), wherein the method further comprises selecting the pool of one or more peptides based on the whole virus or oncoprotein.

[0110] (9) The method of any one of (1)-(8), wherein the method further comprises selecting the one or more peptide pools based on at least one of CD4 T cell interaction or CD8 T cell interaction.

[0111] (10) The method of any one of (1)-(9), wherein the method further comprises selecting the one or more peptide pools based on the average binding prediction of these HLA-I functional groups.

[0112] (11) The method of any one of (1)-(10), wherein the method further comprises selecting the one or more peptide pools based on the average binding prediction of these HLA-II functional groups.

[0113] (12) The method of any one of (1)-(11), wherein the method further comprises selecting the one or more peptide pools based on the regions of predicted binding frequencies in these HLA-I and HLA-II functional groups.

[0114] (13) The method of any one of (1)-(12), wherein the method further comprises selecting the one or more peptide pools based on pan-HLA binding prediction.

[0115] (14) The method of any one of (1)-(13), wherein the T cell response test includes at least one of the following: enzyme-linked immunosorbent assay (ELISpot) assay, cytotoxic T lymphocyte (CTL) assay and DNA barcoded peptide-MHC (pMHC) multimer assay.

[0116] (15) The method of any one of (1)-(14), wherein the T cell response test further comprises testing the T cell response as determined by the synthetic TCR.

[0117] (16) The method as described in (15), wherein the synthetic TCR assay is intended to supplement the T-cell response data of the patient or patient population.

[0118] (17) The method of any one of (1)-(16), wherein the method further comprises using the TCR assay to classify or estimate the patient status.

[0119] (18) The method as described in (17), wherein the method further comprises administering therapeutic treatment to the patient based on the classified or estimated patient status.

[0120] (19) A computer program product comprising a non-transitory computer-readable medium including processor-executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations such that:

[0121] Obtain the first multiple inputs representing various peptides;

[0122] The first multiple inputs are used to train an artificial neural network (ANN) defining a pan-human leukocyte antigen (HLA) binding classifier model, wherein the trained HLA binding classifier model is configured to independently determine the average binding prediction of overlapping peptides at each location of a virus or oncoprotein for each of multiple test HLAs containing HLA-I and HLA-II functional groups.

[0123] Obtain a second multiple input representing a viral or oncoprotein encoded as multiple peptides;

[0124] The second plurality of inputs are fed into the trained HLA binding classifier model, wherein the trained HLA binding classifier is configured to determine the average binding prediction of overlapping peptides among the plurality of peptides.

[0125] Based on these average binding predictions, one or more peptide pools are selected from the multiple peptides;

[0126] Obtain a third multiple input associated with multiple blood samples, where these blood samples represent patients or patient groups;

[0127] A T-cell response model is instantiated based on the one or more peptide pools and the third plurality of inputs; wherein the T-cell response model is trained to predict peptides and protein fragments that are highly likely to elicit a T-cell response based on validated T-cell epitopes and peptides that fail to elicit a T-cell response.

[0128] Responsive T cells identified based on T cell response criteria were sequenced using a sequencer.

[0129] Based on data obtained from sequencing these responsive T cells, detect one or more T cell response patterns common to the patient or patient population;

[0130] Train a TCR classifier / regression model to predict or estimate patient status using a dataset based on one or more T cell response patterns;

[0131] Use a trained TCR classifier / regression model to determine the minimal set of T cell receptors for classifying or estimating the patient's condition; and

[0132] One or more primers are designed based on the identified minimal set of T-cell receptors, which define a TCR assay for classifying or estimating the patient's condition.

[0133] (20) A computer system comprising:

[0134] A memory that stores one or more instructions for designing T-cell receptor (TCR) assays; and

[0135] One or more processors, coupled to the memory, are configured to execute the one or more instructions to perform an operation such that:

[0136] Obtain the first multiple inputs representing various peptides;

[0137] The first multiple inputs are used to train an artificial neural network (ANN) defining a pan-human leukocyte antigen (HLA) binding classifier model, wherein the trained HLA binding classifier model is configured to independently determine the average binding prediction of overlapping peptides at each location of a virus or oncoprotein for each of multiple test HLAs containing HLA-I and HLA-II functional groups.

[0138] Obtain a second multiple input representing a viral or oncoprotein encoded as multiple peptides;

[0139] The second plurality of inputs are fed into the trained HLA binding classifier model, wherein the trained HLA binding classifier is configured to determine the average binding prediction of overlapping peptides among the plurality of peptides.

[0140] Based on these average binding predictions, one or more peptide pools are selected from the multiple peptides;

[0141] Obtain a third multiple input associated with multiple blood samples, where these blood samples represent patients or patient groups;

[0142] A T-cell response model is instantiated based on the one or more peptide pools and the third plurality of inputs; wherein the T-cell response model is trained to predict peptides and protein fragments that are highly likely to elicit a T-cell response based on validated T-cell epitopes and peptides that fail to elicit a T-cell response.

[0143] Responsive T cells identified based on T cell response criteria were sequenced using a sequencer.

[0144] Based on data obtained from sequencing these responsive T cells, detect one or more T cell response patterns common to the patient or patient population;

[0145] Train a TCR classifier / regression model to predict or estimate patient status using a dataset based on one or more T cell response patterns;

[0146] Use a trained TCR classifier / regression model to determine the minimal set of T cell receptors for classifying or estimating the patient's condition; and

[0147] One or more primers are designed based on the identified minimal set of T-cell receptors, which define a TCR assay for classifying or estimating the patient's condition.

[0148] It should be understood that the methods described herein can be implemented using digital processing systems (such as conventional general-purpose computer systems). Alternatively, a special-purpose computer designed or programmed to perform only one function can be used. Figure 5 This is an illustration of an exemplary computing device that can implement the embodiments described herein. Figure 5 The computing device can be used to perform functions designed for TCR determination according to some embodiments. The computing device includes a central processing unit (CPU) 502 coupled to memory 504 and mass storage device 508 via bus 506. Mass storage device 508 represents a persistent data storage device, such as a floppy disk drive or a fixed disk drive, and may be local or remote in some embodiments. In some embodiments, mass storage device 508 may be implemented as backup memory. Memory 504 may include read-only memory, random access memory, etc. In some embodiments, applications residing on the computing device may be stored on or accessed through a computer-readable medium (e.g., memory 504 or mass storage device 508). Applications may also be in the form of modulated electronic signals accessed via a network modem or other network interface of the computing device. It should be understood that in some embodiments, CPU 502 may be embodied in a general-purpose processor, a dedicated processor, or a specially programmed logic device.

[0149] Display 512 communicates with CPU 502, memory 504, and mass storage device 508 via bus 506. Display 512 is configured to display any visualization tools or reports associated with the system described herein. Input / output device 510 is coupled to bus 506 to transmit information from command selection to CPU 502. It should be understood that data to and from external devices can be transmitted via input / output device 510. CPU 502 can be defined to perform the functions described herein to implement the reference... Figure 1-4 The described functionality. In some embodiments, code embodying this functionality may be stored in memory 504 or mass storage device 508 for execution by a processor (such as CPU 502). The operating system on the computing device may be iOS. TM MS-WINDOWS TM OS / 2 TM UNIX TM LINUX TM Or other known operating systems. It should also be understood that the embodiments described herein can be integrated with virtualized computing systems.

[0150] Many details have been set forth in the foregoing description. However, it will be apparent to those skilled in the art that the invention can be practiced without these specific details. In some instances, well-known structures and devices have been shown in block diagram form rather than in detail to avoid obscuring the invention.

[0151] It should be understood that the above description is intended to be illustrative and not restrictive. Many other embodiments will be apparent to those skilled in the art upon reading and understanding the above description. Although the invention has been described with reference to specific exemplary embodiments, it will be recognized that the invention is not limited to the described embodiments but can be practiced with modifications and changes within the spirit and scope of the appended claims. Therefore, the specification and drawings should be regarded as illustrative rather than restrictive. Accordingly, the scope of the invention should be determined by reference to the appended claims and the full scope of the equivalents entitled to such claims.

[0152] It should be understood that although the terms first, second, etc., may be used herein to describe various steps or calculations, these steps or calculations should not be limited by these terms. These terms are used only to distinguish one step or calculation from another. For example, without departing from the scope of this disclosure, a first calculation may be referred to as a second calculation, and similarly, a second step may be referred to as a first step. As used herein, the terms “and / or” and the symbol “I” include any and all combinations of one or more of the listed related items. Unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” are intended to include the plural forms as used herein. It will be further understood that the terms “comprises,” “comprising,” “includes,” and “including,” when used herein, specify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. Therefore, the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be restrictive.

[0153] Although the method operations are described in a specific order, it should be understood that other operations may be performed between the described operations, the described operations may be adjusted so that they occur at slightly different times, or the described operations may be distributed in a system that allows the processing operations to occur at various intervals associated with the processing.

[0154] It should also be noted that in some alternative embodiments, the indicated functions / actions may not occur in the order shown in the figures. For example, two figures shown successively may actually be performed substantially simultaneously or sometimes in reverse order, depending on the functions / actions involved. Considering the above embodiments, it should be understood that the embodiments may employ various computer-implemented operations involving data stored in a computer system. These operations are operations that require physical manipulation of physical quantities. Typically (but not necessarily), these physical quantities exist in the form of electrical or magnetic signals and can be stored, transmitted, combined, compared, and otherwise manipulated. Further, the manipulations performed are generally referred to by terms such as generating, identifying, determining, or comparing. Any operation described herein that forms part of the embodiments is a useful machine operation. These embodiments also relate to an apparatus or device for performing these operations. The device may be specifically constructed for a particular need, or the device may be a general-purpose computer selectively activated or configured by a computer program stored in the computer. In particular, various general-purpose machines may be used with computer programs written in accordance with the teachings herein, or it may be more convenient to construct a more specialized device to perform the required operations.

[0155] Modules, applications, layers, agents, or other method-operable entities can be implemented as hardware, firmware, or a processor executing software, or a combination thereof. It should be understood that, in the software-based embodiments disclosed herein, the software can be embodied in a physical machine (e.g., a controller). For example, a controller may include a first module and a second module. The controller can be configured to perform various actions, such as those of methods, applications, layers, or agents.

[0156] The embodiments can also be implemented as computer-readable code on a non-transitory computer-readable medium. A computer-readable medium is any data storage device that can store data that can subsequently be read by a computer system. Examples of computer-readable media include hard disk drives, network access storage (NAS), read-only memory, random access memory, CD-ROM, CD-R, CD-RW, magnetic tape, flash memory devices, and other optical and non-optical data storage devices. Computer-readable media can also be distributed across network-coupled computer systems, allowing the computer-readable code to be stored and executed in a distributed manner. The embodiments described herein can be practiced with a variety of computer system configurations, including handheld devices, tablet computers, microprocessor systems, microprocessor-based or programmable consumer electronics, microcomputers, mainframe computers, etc. The embodiments can also be practiced in distributed computing environments, where tasks are performed by remote processing devices linked via wired or wireless networks.

[0157] In various embodiments, one or more portions of the methods and mechanisms described herein can form part of a cloud computing environment. In such embodiments, resources can be provided as a service over the Internet according to one or more different models. Such models can include Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). In IaaS, computing infrastructure is delivered as a service. In this case, computing devices are typically owned and operated by the service provider. In the PaaS model, software tools and underlying devices used by developers to develop software solutions can be provided as a service and hosted by the service provider. SaaS typically includes on-demand licensing of Software as a Service by the service provider. The service provider can host the software or deploy it to customers for a given period of time. Many combinations of the above models are possible and conceivable.

[0158] Various units, circuits, or other components may be described or claimed as being “configured to” perform one or more tasks. In this context, the phrase “configured to” is used to imply such a structure by indicating that the unit / circuit / component includes a structure (e.g., a circuit system) that performs one or more tasks during operation. Thus, the unit / circuit / component can be considered to be configured to perform a task even when the specified unit / circuit / component is currently inoperable (e.g., not switched on). Units / circuit / components used with the language “configured to” include hardware; for example, circuits, memory storing program instructions executable to perform operations, etc. The statement that a unit / circuit / component is “configured to” perform one or more tasks is not expressly intended to invoke paragraph 6 of 35U.SC112 for that unit / circuit / component. Additionally, “configured to” can include a general structure (e.g., a general circuit system) that is manipulated by software and / or firmware (e.g., an FPGA or a general-purpose processor executing software) to operate in a manner capable of performing the tasks(s) in question. "Configured to" may also include modifying manufacturing processes (such as semiconductor manufacturing facilities) to manufacture devices (such as integrated circuits) suitable for performing one or more tasks.

Claims

1. A method for designing a T-cell receptor (TCR) assay performed by a processor-based TCR assay module, the method comprising: Obtain the first multiple inputs representing various peptides; The first multiple inputs are used to train an artificial neural network (ANN) defining a pan-human leukocyte antigen (HLA) binding classifier model, wherein the trained HLA binding classifier model is configured to independently determine the average binding prediction of overlapping peptides at each location of a virus or oncoprotein for each of multiple test HLAs containing HLA-I and HLA-II functional groups. Obtain a second multiple input representing a viral or oncoprotein encoded as multiple peptides; The second plurality of inputs are fed into the trained HLA binding classifier model, wherein the trained HLA binding classifier is configured to determine the average binding prediction of overlapping peptides among the plurality of peptides. Based on these average binding predictions, one or more peptide pools are selected from the multiple peptides; Obtain a third multiple input associated with multiple blood samples, where these blood samples represent patients or patient groups; A T-cell response model is instantiated based on the one or more peptide pools and the third plurality of inputs; wherein the T-cell response model is trained to predict peptides and protein fragments that are highly likely to elicit a T-cell response based on validated T-cell epitopes and peptides that fail to elicit a T-cell response. Responsive T cells identified based on T cell response criteria were sequenced using a sequencer. Based on data obtained from sequencing these responsive T cells, detect one or more T cell response patterns common to the patient or patient population; Train a TCR classifier / regression model to predict or estimate patient status using a dataset based on one or more T cell response patterns; Use a trained TCR classifier / regression model to determine the minimum set of T cell receptors used to classify or estimate the patient's condition; as well as One or more primers are designed based on the identified minimal set of T-cell receptors, which define a TCR assay for classifying or estimating the patient's condition.

2. The method of claim 1, wherein training the HLA-based classifier model comprises: Multiple test HLAs encoding variable-length proteins were obtained, wherein the multiple test HLAs comprise HLA-I and HLA-II functional groups; The classifier model is used to process the variable-length peptides encoding the viral protein and the variable-length proteins corresponding to the multiple test HLAs, such that, independently of the test HLAs, the classifier model is operable to determine the average binding prediction of overlapping peptides at each position of the viral protein. Independently based on HLA testing: The aggregated average binding predictions are mapped to locations along the test viral proteins to enable indicator peptide-HLA interactions; Use a sliding window of fixed length to determine the location of the most recent maximum value of these combined average predictions; The region of highest value is determined by selecting the location of the most recent maximum value within the range of the highest percentage of the predicted average. Select peptides that are classified as conjugates that overlap with the region of highest maximum value; as well as Determine the pan-HLA maximum region, wherein the determination includes setting unselected locations to zero, calculating the average along the average combined with the predicted HLA axis, and selecting the pan-HLA maximum value within the highest percentage of the value based on the average. Independently for each of the HLA-I and HLA-II functional groups: Selected peptides classified as conjugates are filtered to identify candidate peptides that overlap with these maximum maxima regions based on a summary of these pan-HLA maximum regions. as well as Incorporate one or more of these candidate peptides into mRNA-based vaccines or therapeutic treatments for patients.

3. The method of any one of claims 1-2, wherein training the TCR classifier / regression model to predict or estimate patient status comprises: Distinguish between patient-state-specific TCR sequences and general TCR sequences associated with patients who do not represent the target disease associated with that patient state; This differentiation is used to identify T cell response patterns; as well as This TCR classifier / regression model is generated based on the identified T cell response patterns.

4. The method of any one of claims 1-3, wherein training the TCR classifier / regression model to predict or estimate patient status comprises: Distinguish between patient-state-specific TCR sequences and general TCR sequences associated with patients who do not represent the target disease associated with that patient state; Based on this distinction, TCR sequences in the sequence embedding space associated with the target disease can be identified; as well as The TCR classifier / regression model is generated using the identified TCR sequences.

5. The method of any one of claims 1-4, wherein determining the minimal set of T cell receptors comprises: Retrieve the predicted scores of multiple patients using a trained TCR classifier / regression model; as well as Select one or more TCR modes based on the retrieved prediction scores.

6. The method of any one of claims 1-5, wherein the method further comprises selecting the one or more peptide pools based on one or more of the following: a specific site, hotspot, or receptor-binding domain of the virus or oncoprotein.

7. The method of any one of claims 1-6, wherein the method further comprises selecting the one or more peptide pools based on multiple regions or hotspots of the virus or oncoprotein.

8. The method of any one of claims 1-7, wherein the method further comprises selecting the one or more peptide pools based on the whole virus or oncoprotein.

9. The method of any one of claims 1-8, wherein the method further comprises selecting the one or more peptide pools based on at least one of CD4 T cell interaction or CD8 T cell interaction.

10. The method of any one of claims 1-9, wherein the method further comprises selecting the one or more peptide pools based on the average binding prediction of these HLA-I functional groups.

11. The method of any one of claims 1-10, wherein the method further comprises selecting the one or more peptide pools based on the average binding prediction of these HLA-II functional groups.

12. The method of any one of claims 1-11, wherein the method further comprises selecting the one or more peptide pools based on regions of predicted binding frequencies in these HLA-I and HLA-II functional groupings.

13. The method of any one of claims 1-12, wherein the method further comprises selecting the one or more peptide pools based on pan-HLA binding prediction.

14. The method of any one of claims 1-13, wherein the T cell response assay comprises at least one of the following: enzyme-linked immunosorbent assay (ELISpot), cytotoxic T lymphocyte (CTL) assay, and DNA barcoded peptide-MHC (pMHC) multimer assay.

15. The method of any one of claims 1-14, wherein the T-cell response assay further comprises testing the T-cell response as determined by the synthetic TCR.

16. The method of claim 15, wherein the synthetic TCR assay is intended to supplement T-cell response data of the patient or patient population.

17. The method of any one of claims 1-16, wherein the method further comprises using the TCR assay to classify or estimate the patient's condition.

18. The method of claim 17, wherein the method further comprises administering therapeutic treatment to the patient based on the classified or estimated patient condition.

19. A computer program product comprising a non-transitory computer-readable medium including processor-executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations such that: Obtain the first multiple inputs representing various peptides; The first multiple inputs are used to train an artificial neural network (ANN) defining a pan-human leukocyte antigen (HLA) binding classifier model, wherein the trained HLA binding classifier model is configured to independently determine the average binding prediction of overlapping peptides at each location of a virus or oncoprotein for each of multiple test HLAs containing HLA-I and HLA-II functional groups. Obtain a second multiple input representing a viral or oncoprotein encoded as multiple peptides; The second plurality of inputs are fed into the trained HLA binding classifier model, wherein the trained HLA binding classifier is configured to determine the average binding prediction of overlapping peptides among the plurality of peptides. Based on these average binding predictions, one or more peptide pools are selected from the multiple peptides; Obtain a third multiple input associated with multiple blood samples, where these blood samples represent patients or patient groups; A T-cell response model is instantiated based on the one or more peptide pools and the third plurality of inputs; wherein the T-cell response model is trained to predict peptides and protein fragments that are highly likely to elicit a T-cell response based on validated T-cell epitopes and peptides that fail to elicit a T-cell response. Responsive T cells identified based on T cell response criteria were sequenced using a sequencer. Based on data obtained from sequencing these responsive T cells, detect one or more T cell response patterns common to the patient or patient population; Train a TCR classifier / regression model to predict or estimate patient status using a dataset based on one or more T cell response patterns; Use a trained TCR classifier / regression model to determine the minimum set of T cell receptors used to classify or estimate the patient's condition; as well as One or more primers are designed based on the identified minimal set of T-cell receptors, which define a TCR assay for classifying or estimating the patient's condition.

20. A computer system comprising: A memory that stores one or more instructions for designing T-cell receptor (TCR) assays; and One or more processors, coupled to the memory, are configured to execute the one or more instructions to perform an operation such that: Obtain the first multiple inputs representing various peptides; The first multiple inputs are used to train an artificial neural network (ANN) defining a pan-human leukocyte antigen (HLA) binding classifier model, wherein the trained HLA binding classifier model is configured to independently determine the average binding prediction of overlapping peptides at each location of a virus or oncoprotein for each of multiple test HLAs containing HLA-I and HLA-II functional groups. Obtain a second multiple input representing a viral or oncoprotein encoded as multiple peptides; The second plurality of inputs are fed into the trained HLA binding classifier model, wherein the trained HLA binding classifier is configured to determine the average binding prediction of overlapping peptides among the plurality of peptides. Based on these average binding predictions, one or more peptide pools are selected from the multiple peptides; Obtain a third multiple input associated with multiple blood samples, where these blood samples represent patients or patient groups; A T-cell response model is instantiated based on the one or more peptide pools and the third plurality of inputs; wherein the T-cell response model is trained to predict peptides and protein fragments that are highly likely to elicit a T-cell response based on validated T-cell epitopes and peptides that fail to elicit a T-cell response. Responsive T cells identified based on T cell response criteria were sequenced using a sequencer. Based on data obtained from sequencing these responsive T cells, detect one or more T cell response patterns common to the patient or patient population; Train a TCR classifier / regression model to predict or estimate patient status using a dataset based on one or more T cell response patterns; Use a trained TCR classifier / regression model to determine the minimum set of T cell receptors used to classify or estimate the patient's condition; as well as One or more primers are designed based on the identified minimal set of T-cell receptors, which define a TCR assay for classifying or estimating the patient's condition.

Images