Methods for profiling immune repertoires

Abstract

This invention disclosure provides nucleic acids, reagents, and methods for profding cell receptor repertoires. The disclosed method is able to distinguish which immune receptor arrangements are present in single cells from amongst the millions of possible immune receptor arrangements. Furthermore, the method can report on the immune receptor arrangement that is present in each of 100,000 or more cells in a single assay, thus defining an individual's immune repertoire. Notably, the disclosed method is applicable for profiling repertoires of various receptors with somatic rearrangements, including that of T-cell receptors or B-cell receptors.

Description

[0001]Docket No.: RU 2023-126 / FR: 096747.00494 METHODS FOR PROFILING IMMUNE REPERTOIRES REFERENCE TO A SEQUENCE LISTING This application contains a Sequence Listing which has been submitted electronically in xml format and is hereby incorporated by reference in its entirety. Said xml copy, created on September 4, 2024, is named “Seq listing (096747.00494).xml” and is 55,416 bytes in size. CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority under 35 U.S.C. §119(e) to U.S. Provisional Patent Application No.63 / 582,589, filed September 14, 2023. The foregoing application is incorporated by reference herein in its entirety. FIELD OF THE INVENTION This invention relates to methods for profiling immune repertoires and more particularly, to profiling T-cell receptor repertoires and B-cell receptor repertoires. This invention also relates to diagnosis of disease conditions in which the immune cell repertoire is altered. BACKGROUND OF THE INVENTION Two major components of the immune system, B cells expressing B-cell receptors (antibodies) and T cells expressing T-cell receptors (TCRs), recognize highly diverse antigens through the creation of a large repertoire of gene sequences that are not present in the germ line that a person inherits. These gene sequences are created during immune cell maturation by a DNA rearrangement process that recombines diverse segments of a gene called variable (V) regions, diversity (D) regions, and joining (J) regions. Because this gene rearrangement process is random, the number of possible V-D-J arrangements is the product of the number of V, D, and J gene segments on the chromosome, which amounts to several million arrangements. This combinatorial diversity is increased further by random nucleotide excisions and additions at the V, D, and J junctions during the recombination process. Millions of clones of T cells, harboring unique combinations of the V, D, and J regions, exist in the blood circulation of healthy individuals. During infections, autoimmune diseases, and certain hematologic cancers, subsets of these clones expand, which leads to a reduction in the diversity of clones in the repertoire. Such alterations in the population structure of the immune cell repertoire can be used as diagnostic hallmarks of these Docket No.: RU 2023-126 / FR: 096747.00494 diseases, and for the management of their therapeutics. Furthermore, a subset of circulating T-cell clones migrate to sites of infection and cancers. Analysis of the T-cell repertoire at such sites is valuable for understanding and managing those diseases. Currently, next-generation sequence analysis (NGS) after polymerase chain reaction (PCR) amplification of the recombined genes is the main method available for the characterization of immune cell repertoires. This technique is expensive, takes a long time to perform, requires special bioinformatic expertise, and suffers from inherent limitations, including differential amplification of some variants at the expense of others. In addition, in normal NGS-based TCR profiling, the information about which pair of α and β receptor chains is used together in the same T cells is not captured, because during cell lysis and library preparation, this information is lost. A number of NGS strategies, including single-cell sequencing, are employed to determine the pairing status of the α and β receptor chains, but those strategies add to the cost of the assays. Thus, a simpler and more rapid method for characterizing immune cell repertoires is needed, without relying on NGS, and without suffering from the limitations associated therewith. SUMMARY OF THE INVENTION This disclosure addresses the above need in a number of aspects. In one aspect, this disclosure provides a method for profiling T-cell or B-cell receptor immune repertoires from multiple T cells or B cells possessing nucleic acid templates that encode chains of T-cell or B-cell receptors. In some embodiments, the method comprises: (a) providing a cell suspension comprising the cells suspended in a solution that comprises amplification and identification reagents, a plurality of primer pairs, a plurality of probes, and at least one polymerase for amplification and identification of a portion of receptor chains of the receptors; (b) fractionating the cell suspension into a plurality of reaction vessels, such that each reaction vessel receives one cell along with the amplification and identification reagents; (c) performing cell lysis and amplification; (d) recording fluorescence signals from each reaction vessel, from each probe as a function of temperature, to generate melting profiles of hybrids formed by amplified targets and the probes; (e) performing an analysis of the melting profiles obtained from each reaction vessel to identify an immune signature in sequences of the receptors that are present in each cell; and (f) performing an aggregated analysis of immune signatures of the cells being analyzed, to determine an immune repertoire of the cells. Docket No.: RU 2023-126 / FR: 096747.00494 In some embodiments, the receptors comprise a T-cell receptor, and wherein the plurality of primer pairs comprise a first primer pair set that is capable of hybridizing to and amplifying a target region in the α chain encompassing both the Vα region and the Jα region, a second primer pair set that is capable of hybridizing to and amplifying a target region in the β chain encompassing both the Vβ region and the Jβ region; and these primer pair sets are chosen such that essentially all possible combinations of Vα-Jα in the α chain and essentially all possible combinations of Vβ- Jβ in the β chain are amplified. In some embodiments, the receptors comprise a T-cell receptor, and wherein the plurality of probes comprise a first probe subset that is capable of hybridizing to a first amplicon in the Vα region and fluoresces in a first color; a second probe subset that is capable of hybridizing to the first amplicon in the Jα region and fluoresces in a second color; a third probe subset that is capable of hybridizing to a second amplicon in the Vβ region and fluoresces in a third color; and a fourth probe subset that is capable of hybridizing to the second amplicon in the Jβ region and fluoresces in a fourth color; and the melting profiles of those hybrids are recorded and analyzed for each fluorescence color. In some embodiments, the cells are B cells and the receptors are B-cell receptors. In some embodiments, the immune signature of each cell is obtained through the use of one or more machine learning models. In some embodiments, the aggregated analysis of the immune signatures of all cells being analyzed is performed to obtain the immune repertoire of the cells. In some embodiments, the method comprises determining population characteristics of the immune repertoire comprising Shannon’s entropy, and identifying the expanded clones and public clones. In some embodiments, one or more characteristics of the immune repertoire are used for disease diagnosis or therapy management. In some embodiments, a different number of probe subsets are used. In some embodiments, the plurality of probes comprise 2 to 25 probes. In some embodiments, the probes comprise sloppy molecular beacon probes. In some embodiments, the length of each probe is between 10 to 50 nucleotides. Docket No.: RU 2023-126 / FR: 096747.00494 In some embodiments, the hybrids have a melting temperature of about 80 °C for probes that do not form mismatches with their target region and of about 25 ºC for probes that form the maximum number of mismatches with their target region. In some embodiments, the plurality of probes comprise probes having an internal stem or an external stem. In some embodiments, the amplification is performed by a polymerase chain reaction, or a recombinase polymerase amplification, or a transcription-mediated amplification, or a strand- displacement amplification. In some embodiments, the nucleic acid templates comprise mRNAs, and the amplification is performed by including a reverse transcriptase in the reactions. In some embodiments, the amplification is asymmetric, producing a larger amount of the strand that is the target of the probes. In some embodiments, cell lysis is performed by heating prior to amplification. In some embodiments, the plurality of reaction vessels are aqueous droplets. In some embodiments, the step of fractionating the cells into the plurality of reaction vessels is performed by emulsification or microfluidics. The foregoing summary is not intended to define every aspect of the disclosure, and additional aspects are described in other sections, such as the following detailed description. The entire document is intended to be related as a unified disclosure, and it should be understood that all combinations of features described herein are contemplated, even if the combination of features are not found together in the same sentence, or paragraph, or section of this document. Other features and advantages of the invention will become apparent from the following detailed description. It should be understood, however, that the detailed description and the specific examples, while indicating specific embodiments of the disclosure, are given by way of illustration only, because various changes and modifications within the spirit and scope of the disclosure will become apparent to those skilled in the art from this detailed description. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a schematic representation of TCR repertoire characterization by single T-cell droplet polymerase chain reaction (PCR) amplification and melting profile analysis utilizing sloppy molecular beacon probes. Docket No.: RU 2023-126 / FR: 096747.00494 Figure 2 is a schematic representation of the organization of the TCR α and TCR β genes in germ line cells before somatic recombination and in the matured T cells after somatic recombination. The approximate locations of the primers used for amplification are also shown. Figure 3 is a schematic representation of the alignment of the variable regions (V) of the TCR β gene segments. A small portion of the alignment that shows high levels of homology between the different gene segments is depicted. The shaded nucleotide residues display at least 75% identity. The target regions of six sloppy molecular beacon probes that can be used to discriminate this set of TCR β gene segments are shown at the bottom of the figure. SEQ ID NOs of the sequences shown in Figure 3 were assigned according to Table 1. Figure 4 shows a hierarchical clustering dendrogram of the TCR βV variable region sequences shown in Figure 3. The sequences within each shaded region are more closely related to each other than the sequences between these clusters. For example, these clusters could be used to design four SMBs. Figure 5 shows the fluorescence intensity of sloppy molecular beacons SMB 1, SMB 2, SMB 3, and SMB 4, all containing two-base pair stem sequences, hybridized to 48 oligonucleotides that represent portions of the Vβ region as a function of temperature (left-hand plots), and the negative derivative of fluorescence intensity with respect to temperature (-dI / dT) as a function of temperature (right-hand plots). Figure 6 shows the experimentally determined melting temperatures of the hybrids formed by the 48 oligonucleotides that are each present in a portion of the Vβ region with sloppy molecular beacon probes SMB 1, SMB 2, SMB 3, and SMB 4, each of which contains a two-base pair stem. Figure 7 shows the correlation between the predicted and experimentally determined melting temperatures of the hybrids formed by the binding of sloppy molecular beacon probes SMB 1, SMB 2, SMB 3, and SMB 4, each of which contain a two-base pair stem, to the 48 oligonucleotides that represent a portion of the Vβ region. This excellent correlation indicates that SMBs can be designed bioinformatically, without extensive trials and testing. Figure 8 shows the melting profile of four sloppy molecular beacon probes, SMB 1, SMB 2, SMB 3, and SMB 4, each of which contains a two-base pair stem, with one target Docket No.: RU 2023-126 / FR: 096747.00494 oligonucleotide (TRBV13), either individually (gray curves), or together in the same reaction (black curve). Figure 9 shows the experimentally determined different melting temperature profiles of the hybrids formed by the four sloppy molecular beacon probes used together in the same reaction (SMB 1, SMB 2, SMB 3, and SMB 4, each of which contain a two-base pair stem) with the 48 oligonucleotides whose sequences are present in portions of the Vβ region. The gray curves show the fluorescence intensity vs. temperature, and the black curves show the negative first derivative of the fluorescence intensity vs. temperature as a function of temperature. Figure 10 shows a “confusion matrix” diagram for a machine-learning algorithm that can detect and distinguish the 48 TCR βV sequences with 96% accuracy through an analysis of the compound melting profiles generated by the four different sloppy molecular beacon probes (SMB 1, SMB 2, SMB 3, and SMB 4, each of which contain a two-base pair stem). The dark gray quadrants along the diagonal signify an accurate prediction, whereas the lighter gray quadrants off of the diagonal represent the very few cases where the algorithm was not able to make an accurate call. “NTC” refers to the no-target controls; and the other identifying entries correspond to the TCR βV sequences that are shown in Table 1. Figure 11 shows a schematic diagram depicting the locations of the primer binding sites and the SMB binding sites over the TCR βV and J regions within the recombined TCR genes in some of the embodiments of this invention. Black areas indicate sites where the sequences have differences between each other, and gray areas indicate sequences that are conserved. The empty white areas in the 5’-encoded tail identify encoded sequence variations in the tail of the primer that become part of the amplified product (and that are then detected by a set of SMBs that are specific to the amplicon containing the encoded tail). Figure 12 shows the fluorescence intensity vs. temperature profiles of sloppy molecular beacon probe SMB 3 containing either a 6, 4, 2, or 0 base pair stem sequence hybridized to 48 oligonucleotides that represent portions of the Vβ region (plots on the left-hand side of the figure). The negative first derivative of the fluorescence vs. temperature curves (-dI / dT) as a function of temperature is shown on the right-hand side of the figure. Figure 13 shows the melting temperatures of sloppy molecular beacon probe SMB3, containing either a 6, 4, 2, or 0 base pair stem sequence hybridized to 48 oligonucleotides that Docket No.: RU 2023-126 / FR: 096747.00494 represent portions of the Vβ region. This data is derived from the negative first derivatives shown in Figure 12. The symbols located on the X axis are the few “dropout” cases for which no melting transition was observed. Figures 14A and 14B show a method of obtaining melt profiles from numerous droplets simultaneously. Figure 14A shows the fluorescence image of a mixture of droplets at 35 °C where individual droplets contain Sloppy Molecular Beacon 1, either alone or in the presence of a TCR target chosen from Table 1. Figure 14B shows the fluorescence of single droplets as a function of temperature. The melt profiles in the graph on the right are sorted and classified using a machine learning algorithm, whereas in the graph on the left, they are presented without any sorting or classification. DETAILED DESCRIPTION OF THE INVENTION This disclosure provides nucleic acids, reagents, and methods for profiling immune repertoires. The disclosed method is able to distinguish which immune receptor arrangements are present in single cells from amongst the millions of possible immune receptor arrangements. Furthermore, the method can report on the immune receptor arrangement that is present in each of 100,000 or more cells in a single assay, thus defining an individual’s immune repertoire. Notably, the disclosed method is applicable for profiling repertoires of various receptors with somatic rearrangements, including that of T-cell receptors (TCRs) or B-cell receptors (BCRs). Methods For Profiling Cell Receptor Repertoires Accordingly, in one aspect, this disclosure provides a method for profiling T-cell or B-cell receptor immune repertoires from multiple T cells or B cells possessing nucleic acid templates that encode chains of T-cell or B-cell receptors. In some embodiments, the method comprises: (a) providing a cell suspension comprising the cells suspended in a solution that comprises amplification and identification reagents, a plurality of primer pairs, a plurality of probes, and at least one polymerase for amplification and identification of a portion of receptor chains of the receptors; (b) fractionating the cell suspension into a plurality of reaction vessels, such that each reaction vessel receives one cell along with the amplification and identification reagents; (c) performing cell lysis and amplification; (d) recording fluorescence signals from each reaction vessel, from each probe as a function of temperature, to generate melting profiles of hybrids formed by amplified targets and the probes; (e) performing an analysis of the melting profiles Docket No.: RU 2023-126 / FR: 096747.00494 obtained from each reaction vessel to identify an immune signature in sequences of the receptors that are present in each cell; and (f) performing an aggregated analysis of immune signatures of the cells being analyzed, to determine an immune repertoire of the cells. An example process for profiling a T-cell receptor (TCR) repertoire is illustrated in Figure 1. The organization of variable regions that are recombined to create functional genes encoding a TCR is depicted in Figure 2. Each TCR is created from two peptide chains, α and β chains, that are encoded by two separate genes. During T-cell maturation, one gene segment from each of the diverse V, D, and J regions is randomly picked and joined together in both the TCR α gene and the TCR β gene. The method as disclosed herein for characterizing the TCR repertoire for the α and β chains involves distributing thousands of single T cells into physical partitions, such as droplets suspended in oil, where each partition contains just a single cell. The recombined TCR genes in these cells, or the mRNAs encoded by them, are then amplified via PCR, utilizing a common set of amplification primers in the presence of a set of sloppy molecular beacon probes (SMBs). Each partition produces two types of amplicons, one derived from the variant of the α TCR and another from the variant of the β TCR that each T cell harbors. Normally due to an allelic exclusion process, only one of the two alleles for the α and β TCR genes undergoes this rearrangement process; however, in rare cases, both alleles are recombined. However, the disclosed method is capable of detecting such cases as well. After PCR amplification, the melting profiles of the SMBs are determined in each partition (Figure 1). Since each of the four variable regions, Vα, Vβ, Jα, and Jβ, is targeted by separate sets of SMBs labeled with different fluorescent colors, each partition will produce four melting profiles in each of four differently colored fluorescence channels. These profiles are then analyzed with a machine learning algorithm to identify each unique TCR arrangement for both the α and β genes. The unique set of four melting profiles in the four fluorescence channels then serves as a composite “immune signature” for a particular TCR pair of α and β gene arrangements in a cell. Simultaneous determination of: (a) the identity of the unique TCR signature in each of the thousands of individual T cells from a single patient sample; and (b) quantitation of the number of cells that contain each different TCR signature, thereby providing an accurate snapshot of that patient’s immune state, allowing for disease diagnosis and therapy management. This analysis is referred to as an “aggregate analysis.” These measurements provide an estimation of the T-cell repertoire diversity, as well as the frequency of each T-cell clone in that patient’s repertoire. Docket No.: RU 2023-126 / FR: 096747.00494 An advantageous aspect of amplification from single cells in isolated partitions, followed by melting profile analysis, is that since competition between diverse amplicons does not occur, amplification bias is mitigated to a great extent. This enables the accurate identification and quantitation of the TCR genes in the population of these cells. A second advantageous aspect of the disclosed method is that information about which pair of α and β receptor chains are present in the same T cell is inherently provided by the process. This is not the case for next generation sequencing (NGS) strategies, where during cell lysis and library preparation the information about α and β pairing is lost, unless elaborate steps, such as single-cell sequencing, are taken to obtain this information. In some embodiments, the analysis of the melting profiles is performed by one or more machine learning models. These models include a neural network, a convolutional neural network (CNN), a deep convolutional neural network (DCNN), a cascaded deep convolutional neural network, a simplified CNN, a shallow CNN, or a combination thereof. In some other embodiments, the melting profile analysis is performed by a classifier such as a Subspace Discriminator classifier, or a classifier based on linear regression, logistic regression, decision trees, support vector machines (SVM), naive Bayes, k-nearest neighbors (k- NN), K-means clustering, random forest, dimensionality reduction algorithms, gradient boosting algorithms, or neural networks. In some embodiments, the full melting profile is utilized, and in other embodiments the melting temperatures are determined from these profiles, and the combination of these melting temperatures is analyzed to determine the immune signature. In these melting profiles, there may be multiple melting transitions stemming from different probes. In some embodiments, the melting temperature profile comprises a thermal denaturation profile (e.g., melting curves) of at least one of the plurality of probe-amplicon hybrids. In some embodiments, the melting temperature is determined by calculating a negative first derivative of melt curves (e.g., plots of temperature versus fluorescence intensity as a function of temperature). In some embodiments, the negative first derivative of the melt curves may be determined during the denaturation segments for several cycles of PCR. In some embodiments, the negative first derivative of the melt curves may be generated by continuously monitoring the PCR reaction. Docket No.: RU 2023-126 / FR: 096747.00494 Partitions The method includes the partitioning of cells into individual partitions, such that generally each partition receives just one cell. Depending on how partitioning is achieved, in some partitions there will be no cell, and in a few other cells there can be more than one cell. One method for partitioning is based on the Poisson distribution of the cells into individual partitions. In another method of partitioning, cells can explicitly be incorporated into partitions in which each partition receives a single cell. As used herein, the term “partition” refers to a volume of fluid (e.g., liquid) that is a separated portion of a bulk volume. A bulk volume may be partitioned into any suitable number (e.g., 102...103...104...105...106...107, etc.) of smaller volumes (i.e., partitions). Partitions may be separated by a physical barrier or by physical forces (e.g., surface tension, hydrophobic repulsion, etc.). Partitions generated from the larger volume may be substantially uniform in size (i.e., monodisperse) or may have non-uniform sizes (i.e., polydisperse). Partitions may be produced by any suitable manner (e.g., emulsion, microfluidics, microspray, etc.). In some embodiments, the partitions are aqueous droplets suspended in an oil matrix, and in other embodiments, the partitions are microwells. As used herein, the term “droplet” refers to a small volume of liquid that is immiscible with its surroundings (e.g., gases, liquids, surfaces, etc.). A droplet may reside upon a surface that is encapsulated by a fluid with which it is immiscible (e.g., the continuous phase of an emulsion). A droplet is typically spherical or substantially spherical in shape, but may be non-spherical. A droplet may be a “simple droplet” or a “compound droplet,” wherein one droplet encapsulates one or more additional smaller droplets. The volume of a droplet and / or the average volume of a set of droplets provided herein may be less than about one hundred microliters. The diameter of a droplet and / or the average diameter of a set of droplets provided herein is typically less than about one millimeter. Droplets may be formed by any suitable technique (e.g., emulsification, microfluidics, injection, etc.), and those droplets may be monodisperse (e.g., substantially monodisperse) or polydisperse. In some embodiments, partitioning the cells into individual partitions may be performed by emulsification or microfluidics. In some embodiments, the partitioning of the cells may be performed by a droplet microactuator, an emulsification device, a microfluidics platform, a Docket No.: RU 2023-126 / FR: 096747.00494 continuous-flow microfluidics device, an array micro-chamber, a microspray device, one or more microplate wells, a reagent immobilization device, an electrical manipulation device, a thermal manipulation device, a mechanical manipulation device, an optical manipulation device, a chemical manipulation device, or combinations thereof. In some embodiments, after the step of partitioning the cells, the cells may be lysed in each partition by heating, such as at about 92 ºC. The initial denaturation step renders the target nucleic acids accessible for amplification. The time devoted to this step can be increased to ensure efficient lysis. To help facilitate this process, mild detergents or other lysis reagents, that do not negatively impact amplification can be included in the partitions. The other lysis reagents include proteinase K. The lysis may or may not be proformed at the same time as amplification. Amplification In some embodiments, the step of amplification is performed by a PCR, recombinase polymerase amplification, transcription-mediated amplification, or strand-displacement amplification. Asymmetric amplification is preferred because it provides better opportunity for probes to bind. In some embodiments, amplification is performed by creating cDNA copies from the mRNAs, and in other embodiments, the DNA is directly amplified. As used herein, the term “asymmetric PCR” refers to the preferential PCR amplification of one strand of a DNA target by adjusting the molar concentration of the primers in a primer pair so that they are unequal. An asymmetric PCR assay produces a predominantly single-stranded product and a smaller quantity of a double-stranded product as a result of the unequal primer concentrations. As asymmetric PCR proceeds, the lower concentration primer is quantitatively incorporated into a double-stranded DNA amplicon, but the higher concentration primer continues to prime DNA synthesis, resulting in continued accumulation of a single-stranded product. Amplification may also include isothermal amplification. The term “isothermal” means conducting a reaction at a substantially constant temperature, i.e., without varying the reaction temperature in which a nucleic acid polymerization reaction occurs. Docket No.: RU 2023-126 / FR: 096747.00494 Target Regions in the Cell Receptor Genes and Probes for Their Identification In some embodiments, the cells that are analyzed are T cells or B cells, and the receptors are T-cell receptors or B-cell receptors. A T-cell receptor (TCR) refers to a receptor located on the membrane of the T cell that recognizes an antigen. TCRs play a central role in the immune system. TCRs have an α chain, a β chain, a γ chain, and a δ chain, with which an αβ or γδ dimer is constituted. TCRs consisting of the combination of the former are called αβ TCRs, and TCRs consisting of the combination of the latter are called γδ TCRs. T cells having such TCRs are respectively called αβ T cells and γδ T cells. The TCRs are structurally very similar to a frequent antibody binding region (Fab fragment) of an antibody produced by B cells, and they recognize antigen molecules in concert with a major histocompatibility (MHC) molecule. Since the TCR gene of each mature T cell has undergone gene rearrangement, an individual person has highly diverse TCRs that enable recognition of a very large number of different antigens Each TCR chain is comprised of a variable domain (V) and a constant domain (C). A constant domain has a short cytoplasm section penetrating the cell membrane. A variable domain is present outside the cell and binds to an antigen-MHC complex. A variable domain has three hypervariable domains or regions called complementarity- determining regions (CDRs), which bind to an antigen-MHC complex. The three CDRs are called CDR1, CDR2, and CDR3. The gene rearrangements in TCRs are similar to the gene rearrangements in B-cell receptors, which are known as immunoglobulins. The organization of the B-cell receptor (BCR) on B cells (antibodies) is similar to the organization of the T-cell receptor on T cells. The diversity in the BCR is created by a gene rearrangement process that is very similar to the rearrangement process in the TCR. The methods described herein in detail for TCR profiling are equally applicable for BCR profiling. In some embodiments, each of the plurality of primer pairs hybridize to and amplify different target regions in the nucleic acid sequence of a cell receptor (e.g., TCR or BCR). In some embodiments, the nucleic acid molecule encoding the cell receptor comprises a cell receptor gene or an mRNA. In some embodiments, the nucleic acid molecule encoding the cell receptor is a cDNA of the mRNA encoding the cell receptor. Docket No.: RU 2023-126 / FR: 096747.00494 As used herein, a “target region,” “target nucleic acid sequence,” or “target sequence” refers to a specific sequence that may include all or part of a longer sequence that needs to be identified. In some embodiments, the target region comprises one or more regions in a V segment and / or a J segment of the cell receptor (e.g., TCR or BCR). In some embodiments, the target region comprises one or more regions in the α chain and / or the β chain of a T-cell receptor. In some embodiments, one or more regions are located in the Vα segment, the Vβ segment, the Jα segment, and / or the Jβ segment of the T-cell receptor. In some embodiments, one or more regions are located in the complementarity determining region 3 (CDR3) of the T-cell receptor. In some embodiments, a set of probes are employed that consist of multiple subsets of probes, where different subsets fluoresce in different fluorescent colors. Each subset may contain from 1 to 10 different probes. In some embodiments, fewer or more probe subsets may be employed. In some embodiments, the cell receptor is a TCR, and its α chain is amplified with one set of primers and its β chain is amplified with a second set of primers generating an α amplicon and a β amplicon, respectively. The primers are designed such that each of these amplicons contains the V and J regions present in the respective α or β genes present in the cell which is being analyzed. The V region of the α amplicon is detected using one subset of probes labeled in one color, and the J region of the α amplicon is detected using a second subset of probes labeled with a second color. Furthermore, the V region of the β amplicon is detected using a third subset of probes labeled in a third color, and the J region of the β amplicon is detected using a fourth subset of probes labeled with a fourth color. Probes As used herein, the term “probe” refers to an oligonucleotide capable of binding to a target nucleic acid having a complementary sequence. Probes may bind target sequences lacking complete complementarity with the probe sequence depending upon the stringency of the hybridization conditions. There may be any number of base pair mismatches, which will interfere with hybridization between the target sequence and the single-stranded nucleic acids described herein. However, if the number of mismatches is so great that no hybridization can occur under Docket No.: RU 2023-126 / FR: 096747.00494 even the least stringent of hybridization conditions, the probe is not a complementary target sequence. The probes according to this disclosure are self-reporting probes. Their fluorescence changes upon binding to its target. These probes can be made in various detection formats, such as dual labeled probes, including linear probes, TaqMan probes, molecular beacon probes, and sloppy molecular beacon (SMB) probes. A “sloppy” probe refers to a probe that is mismatch- tolerant. Mismatch-tolerant probes may hybridize with more than one target sequence to generate a detectable signal at a detection temperature in an assay, and various hybrids so formed will have different melting temperatures. Examples of such probes are hairpin or linear probes with a fluorescent moiety whose level of fluorescence increases upon. See, e.g., U.S. Pat. Nos.5,925,517 and 7,662,550. In some embodiments, the sloppy probes are dual-labeled hairpin probes or molecular beacon probes, described in U.S. Pat. Nos.5,925,517 and 7,662,550. These hairpin probes contain a target binding sequence flanked by a pair of arms complementary to one another. They can be DNA, RNA, or PNA, or a combination of all three types of nucleotides. Furthermore, they can contain modified nucleotides and modified internucleotide linkages. They can have a first fluorophore on one arm and a second fluorophore on the other arm, wherein the absorption spectrum of the second fluorophore substantially overlaps the emission spectrum of the first fluorophore. Such hairpin probes may be “molecular beacon probes” that have a fluorophore on one arm and a quencher on the other arm such that the probes are dark when free in solution. They can also be wavelength-shifting molecular beacon probes with, for example, multiple fluorophores on one arm that interact by fluorescence resonance energy transfer (FRET), and a quencher on the other arm. They can also have a first fluorophore on one arm and second fluorophore on the other arm with an internal quencher molecule in the target binding sequence region. The target binding sequences can be10 to 100 nucleotides in length, preferably 10 to 50 nucleotides in length. Each arm is 2 to 10 nucleotides in length. A portion or the entirety of the arm sequences of the probe can have complementarity to the target sequence, and a portion of the target region of the hairpin probe can participate in forming the hairpin along with one or both arm sequences. Probes according to this disclosure need not possess complementary arms. Docket No.: RU 2023-126 / FR: 096747.00494 By virtue of their ability to bind to more than one variants of a given target sequence, sloppy molecular beacons can be used in assays to detect the presence of one variant of a nucleic acid sequence segment of interest from among a number of possible variants or even to detect the presence of two or more variants. These probes can therefore be used in combinations of two or more in the same assay. Because they differ in target binding sequence, their relative affinities for different variants are different. For example, a first probe may bind strongly to a wild-type sequence, moderately to a first allele, weakly to a second allele and not at all to a third allele; while a second probe may bind weakly to the wild-type sequence and the first variant, and moderately to the second variant and the third variant. Additional sloppy molecular beacon probes will exhibit yet different binding patterns due to their different target binding sequences. Thus, the patterns of the fluorescence emission spectra from combinations of sloppy molecular beacon probes can define different related, but different, sequences. Sloppy molecular beacon probes have the ability to bind to a number of different related target sequences at a low temperature, where they fluoresce brightly. However, as the temperature is raised, the different hybrids formed by the sloppy molecular beacon dissociate at different temperatures, depending on the degree of complementarity, resulting in a reduction of fluorescence intensity. The fewer the number of mismatches in the hybrid, the higher the melting temperature is. By simultaneously utilizing a number of different sloppy molecular beacon probes, each labeled with the same colored fluorophore, and each designed to hybridize to the same target region, it is possible to obtain a high-content melting profile that uniquely identifies that target. Because each of these different probes forms hybrids containing a different number of mismatches, there is a series of different melting transitions, which, when combined, produce the unique high- content identifying profile. In addition, when the same sloppy molecular beacon probe set described above is used against a different, but related, target sequence, the resulting melting profile is distinguishably different. The same set of sloppy molecular beacon probes, all labeled in the same color, can produce dozens of distinctly different melting profiles, each of which identifies a different target. By simultaneously using a number of different sloppy molecular beacon probe subsets, where the probes in each subset are labeled in a different fluorescent color, the melting profiles Docket No.: RU 2023-126 / FR: 096747.00494 obtained in each fluorescent color will independently identify different targets in each of a number of different groups, where each different group is targeted by a differently colored the probe set. These multiplex assays have the ability to distinguish a vast number of distinctive targets in the same assay. For instance, an assay simultaneously utilizing four differently colored subsets of sloppy molecular beacon probes can distinguish as many as 3,000,000 different immune signatures that can be present in T cells. In some embodiments, the plurality of probe-amplicon hybrids have a melting temperature of about 80°C for probes that do not form mismatches with the target region and at about 25 ºC for probes that form the maximum number of mismatches with the target region. In some embodiments, the probes may have an internal stem, and in other embodiments, the probes may have a terminal stem. The stem length can range from 2 to 10 base pairs. In some embodiments, adjacent probes that bind to the same target near each other and then interact via fluorescence resonance energy transfer (FRET) or contact-mediated quenching can also be used in the disclosed method. Examples of such probes may include Lights-On / Lights- Off probes described in M. G. Whitfield et al.2020 Scientific Reports 10, Article 4234. In some embodiments, the set of probes comprises a probe having a chemical modification, such as a modified nucleotide, a modified chemical backbone, or a combination thereof. As used herein, a “chemical modification” refers to a chemical difference in a compound when compared to a naturally occurring counterpart. Chemical modifications of oligonucleotides may include nucleoside modifications (such as sugar moiety modifications and nucleobase modifications) and internucleoside linkage modifications. As used herein, an “internucleoside linkage” refers to a covalent linkage between adjacent nucleosides in an oligonucleotide. In reference to an oligonucleotide, a chemical modification does not include differences only in the nucleobase sequence. Examples of modifications may include, without limitation, a 2’-O-methyl-modified sugar moiety. Probes may consist of ribonucleotides instead of deoxyribonucleotides. Probes may also contain both ribonucleotides and deoxyribonucleotides. In some embodiments, the probes have a fluorophore and a quencher. The probes may have other labels as described in U.S. Pat. No.5,925,517. As used herein, a “fluorophore” includes a molecule that is capable of absorbing energy at a wavelength range and releasing energy at a wavelength range other than the absorbance range. In some embodiments, the fluorophore is a Docket No.: RU 2023-126 / FR: 096747.00494 molecule that is capable of absorbing energy at about 250 nm to about 900 nm, and can release energy at a wavelength range of about 260 nm to about 910 nm. The term “excitation wavelength” refers to the range of wavelengths at which a fluorophore absorbs energy. The term “emission wavelength” refers to the range of wavelengths that the fluorophore releases energy or fluoresces. Some examples of fluorophores and quenchers are found in U.S. Pat. Nos. 5,925,517 and 7,662,550, as well as in Marras et al.2002 Nucleic Acids Research 30, e122., Primers As used herein, the term “primer” refers to any nucleic acid that is capable of specifically hybridizing to a complementary nucleic acid molecule and that provides a free 3’-hydroxyl terminus that can be extended by a nucleic acid polymerase. As used herein, amplification primers are a pair of nucleic acid molecules that can anneal to 5’ or 3’ regions of a gene (plus and minus strands, respectively, or vice-versa) and contain a short region in between. Under appropriate conditions and with appropriate reagents, such primers enable the amplification of a nucleic acid molecule having the nucleotide sequence flanked by the primers. When the cells being analyzed are T cells and the receptor being analyzed is the T cell receptor, primers are designed such that the forward primers will initiate from a portion of the V regions of the a and b chains, and the reverse primers will initiate from a portion of the J regions of those chains. The design of the primers should be such that the region at the junction of the V and J regions is amplified, and the portions of the V and J regions that are amplified are sufficiently long to include the regions targeted by the detection probes. A further criterion for the design of the primers is that all possible combinations of V and J should be amplified without bias. Persons skilled in the art will appreciate that the same primers will not amplify non-recombined cell receptor genes, because the distance between the two primer binding sites will be so large that efficient amplification does not occur. These criteria may be satisfied by one pair of primers for each of the a and b chains, or the criteria may be satisfied by using multiple forward or reverse primers for the a and b chains. An example of a suitable primer set has been described by Robins et al.2009 Blood 114, 4099-4107. Other suitable primer sets are discussed by Ch’ng et al.2019 European Journal of Immunology 49, 1186-1199. Docket No.: RU 2023-126 / FR: 096747.00494 To reduce the number of primers that will satisfy the criteria above, it may be useful to include degenerate bases in the primers so that they will be able to bind to and amplify many possible V and J sequences. Methods for Diagnosis Based on Profiling Cell Receptor Repertoires In some embodiments, the immune signatures of single cells are used to obtain immune repertoires of collections of cells. Immune signatures are sets of multiple melting profiles that, in the case of TCR, are obtained from single cells for the Vα, Jα, Vβ, and Jβ regions, that enable the gene segments that are used in a single cell to be identified. The sets of immune signatures obtained from a large number of single cells in a sample from a person collectively represent the immune repertoire of the cells in that sample. Immune repertoires are a list of the immune signatures and their frequencies in the sample. Since each T-cell clone possesses a unique TCR arrangement, it will have a unique immune signature. The immune repertoires are used for diagnostics and therapy management. In some embodiments, where the T-cell receptors or B-cell receptors are analyzed by the procedures of this invention, it is useful to correlate the characteristics of the immune repertoire with the repertoires from patients confirmed to have certain disease conditions. These correlations may be used to determine diagnosis, prognosis, or therapy management of patients’ disease conditions or disorders. Examples of diseases or disorders may include, without limitation, a malignant disease (e.g., cancer), a viral disease (e.g., HIV-1 infection), a bacterial disease (e.g., tuberculosis), a fungal disease, a protozoan disease, a parasitic disease, an allergic disease, or an autoimmune disease. In some embodiments, where the T-cell receptors or B-cell receptors are analyzed by the procedures of this invention, the immune signatures are used directly to obtain the immune repertoire of a collection of cells. In other embodiments, the immune signatures are first used to determine the Vα, Jα, Vβ, and Jβ sequences, and those sequences are then used to obtain the immune repertoire. In some embodiments, one or more characteristics of the immune repertoire of a subject is compared to the immune repertoire of reference samples obtained from other subjects having a certain disease condition. The subject can be a human or an animal. Docket No.: RU 2023-126 / FR: 096747.00494 As used herein, a “sample” refers to any biological fluid or tissue obtained from a patient that contains cells whose receptor arrangement needs to be determined. The sample may be a “clinical sample,” which is a sample derived from a subject, such as a human patient or veterinary subject. The samples may also include sections of tissues, such as frozen sections taken for histological purposes. Compositions and Kits Described herein are compositions and kits for obtaining immune signatures from cells, for determining immune repertoires from a collection of cells, and for performing clinical diagnostic assays. These compositions and kits include instructions and reagents needed to perform the assays. The reagents include primer sets, probe sets, and amplification reagents, including polymerases, nucleotides, and buffers. The reagents may also include oils and detergents for droplet generation, reagents for cell lysis. These reagents can be provided separately or in various combined mixtures. Additional Definitions Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. As used herein, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. As used herein, the terms “including,” “comprising,” “containing,” or “having” and variations thereof are meant to encompass the items listed thereafter and equivalents thereof as well as additional subject matter unless otherwise noted. As used herein, the phrases “in one embodiment,” “in various embodiments,” “in some embodiments,” and the like are used repeatedly. Such phrases do not necessarily refer to the same embodiment, but they may unless the context dictates otherwise. As used herein, the terms “and / or” or “ / ” means any one of the items, any combination of the items, or all of the items with which is associated. Docket No.: RU 2023-126 / FR: 096747.00494 As used herein, the word “essentially” or “substantially” does not exclude “completely,” e.g., a composition that is “substantially free” from Y may be completely free from Y. Where necessary, the word “substantially” may be omitted from the definition of this disclosure. As used herein, the term “each,” when used in reference to a collection of items, is intended to identify an individual item in the collection but does not necessarily refer to every item in the collection. Exceptions can occur if explicit disclosure or context clearly dictates otherwise. As used herein, the term “approximately” or “about,” as applied to one or more values of interest, refers to a value that is similar to a stated reference value. In some embodiments, the term “approximately” or “about” refers to a range of values that fall within 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less in either direction (greater than or less than) of the stated reference value unless otherwise stated or otherwise evident from the context (except where such number would exceed 100% of a possible value). Unless indicated otherwise herein, the term “about” is intended to include values, e.g., weight percents, proximate to the recited range that are equivalent in terms of the functionality of the individual ingredient, the composition, or the embodiment. As disclosed herein, a number of ranges of values are provided. It is understood that each intervening value, to the tenth of the unit of the lower limit, unless the context clearly dictates otherwise, between the upper and lower limits of that range is also specifically disclosed. Each smaller range between any stated value or intervening value in a stated range and any other stated or intervening value in that stated range is encompassed within the disclosure. The upper and lower limits of these smaller ranges may independently be included or excluded in the range, and each range where either, neither, or both limits are included in the smaller ranges is also encompassed within the invention, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the invention. All methods described herein are performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. In regard to any of the methods provided, the steps of the method may occur simultaneously or sequentially. When the steps of the method occur sequentially, the steps may occur in any order, unless noted otherwise. In cases in which a method comprises a combination of steps, each and every combination or sub- Docket No.: RU 2023-126 / FR: 096747.00494 combination of the steps is encompassed within the scope of the disclosure, unless otherwise noted herein. Each publication, patent application, patent, and other reference cited herein is incorporated by reference in its entirety to the extent that it is not inconsistent with the present disclosure. Publications disclosed herein are provided solely for their disclosure prior to the filing date of the present invention. Nothing herein is to be construed as an admission that the present invention is not entitled to antedate such publication by virtue of prior invention. Further, the dates of publication provided may be different from the actual publication dates, which may need to be independently confirmed. It is understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this application and scope of the appended claims. Examples Example 1. An Exemplary Process for Obtaining T-cell Repertoires from T cells Aligning and Choosing Target Sequences The first step was to download all V, D, and J sequences present in the human T-cell receptor α and β chains from a gene sequence repository. An example of such a repository is Ensembl (Cunningham et al.2022 Nucleic Acids Research 50, D988-995). The sequences within each group were then aligned using a multi-sequence alignment tool. Examples of such multi- sequence alignment tools are MAFFT (Rozewicki et al.2019 Nucleic Acids Research 47, W5-10) and Clustal Omega (Sievers et al. 2011 Molecular Systems Biology 7, 539). In the alignments that these computer programs produce, the most conserved residues were identified. Using these conserved residues, subregions of the sequence that have relatively higher levels of conservation were identified. An example of such a region is shown in Figure 3 for one of the several conserved regions that are present within the TCR Vβ region. In this alignment, the TCR sequence names are shown on the left, followed by their nucleotide sequences on the right. The shaded nucleotides are conserved in 75% or more within this set of TCR Vβ sequences. In order to choose the target regions that will be amplified and probed, a sequence region that displays relatively high levels of Docket No.: RU 2023-126 / FR: 096747.00494 conservation should be chosen, but the sequences that occur in this region should not be perfectly conserved. Following similar procedures, suitable regions of high homology can be identified in the J regions of TCR β, and in the V and J regions of TCR α. This approach is also effective in the case of B-cell receptors analysis, where a similar gene segment rearrangement process gives rise to sequence diversity. Sloppy Molecular Beacon Probes Sloppy molecular beacons (SMBs) are internally quenched probes that can tolerate one or more mismatches while binding to their target. When these probes bind to a fully or partially complementary target, their fluorescence is increased. They contain a fluorophore and a quencher moiety, for example, at their terminals. However, it is possible to attach these labels at internal locations and obtain good background-to-signal ratios, i.e., the fluorescence intensity displayed upon binding to the target divided by the fluorescence intensity before binding. Their probe lengths may vary from 10 to 100 nucleotides, preferably 10 to 40 nucleotides, and their stem lengths may vary from 2 to 10 nucleotides. Under certain situations, it may be possible to design probes that do not have any stem. Such situations may include when the target itself has internal stems, which results in the probe sequence also having internal stems. Such probes do not possess a designed stem, but rather an internal stem serves to bring the fluorophore and the quencher close to each other in the unbound form of the probe. Another situation where a designed stem may not be necessary is when the fluorophore and the quencher have a high affinity for each other, and they tend to bind to each other when present in the same molecule. For the purpose of this application, sloppy molecular beacons may or may not possess a terminal stem. Under certain conditions, it may be advantageous to use adjacent probes that bind to the same target near each other and then interact via fluorescence resonance energy transfer (FRET) or contact-mediated quenching. A useful embodiment of this approach is the use of Lights- On / Lights-Off probes (Whitfield et al.2020 Scientific Reports 10, Article 4234). SMBs can consist of natural nucleotides with a natural phosphodiester backbone, or they can contain modified nucleotides and / or modified chemical backbones. The modified nucleotides may be designed to increase or decrease the specificity of complementary base recognition. Docket No.: RU 2023-126 / FR: 096747.00494 Designing Sloppy Molecular Beacon Probes for T-cell Receptor Regions To discriminate all of the targets in a given region of variability, such as the V region of TCR β discussed above, a set of 2 to 8 SMBs were designed using computer programs. A key consideration in the design of SMBs is that each SMB should bind to a large fraction of the targets in the target set, and that the strengths of their binding should vary from target to target and not be the same. The presence of more mismatches results in weaker probe-target hybrids, and the presence of a lesser number of mismatches results in stronger probe-target hybrids. A second consideration in the choice of the sequences present in the set of SMBs is that those sequences in the target region that do not bind well to one SMB in the set should bind well to one or more other SMBs in the set. In fact, the set of all SMBs that are chosen should be such that all targets are covered by one or more SMB. Although an SMB can have a target that is fully complementary to it, that is not a necessity. Indeed, SMBs can be chosen that are not fully complementary to any target sequence within the target region. These objectives are achieved by choosing appropriate lengths of the target binding region of the SMBs, and then by choosing appropriate locations within the targets to which the SMBs bind. Although the length of the SMB’s target-binding region can vary from 10 to 50 nucleotides, a region was selected that yields probe-target hybrids melting at about 80 ºC for probes that do not form mismatches, and at about 25 ºC with targets whose probes form the maximum number of mismatches in the group of target sequences. The prediction of the melting temperature of each SMB with each target is determined with the aid of computer programs, such as Primer3 (Untergasser et al. 2012 Nucleic Acids Research 40, e115) and the DINAMelt web server (Markham and M. Zuker 2005 Nucleic Acids Research 33, W577-581). After identifying the target binding region for each sloppy molecular beacon probe, the sequence of the reverse complement of its target sequence (including mismatched nucleotides) was selected as the loop of the SMB. These “probe” sequences were then appended to mutually complementary stem sequences that can each vary from 2 to 10 nucleotides in length. Short stems melt at low temperatures and present little impediment for binding of the SMB to the target, while longer stems melt at higher temperatures and present a stronger impediment for binding to the target. Docket No.: RU 2023-126 / FR: 096747.00494 The manner in which stem length and stem strength impact the function of SMBs, and their coverage of a full set of target sequences, is discussed in Example 3. In general, longer stems create impediments to binding to targets, especially for SMBs forming many mismatches with a target sequence, yet small stems enable binding to such targets. Consequently, SMBs with longer stems have higher “dropouts,” i.e., targets that do not produce a characteristic melting transition than SMBs possessing smaller stems (Figures 12 and 13). Following this procedure, “candidate” SMBs were designed bioinformatically for every target in the group of target sequences. After designing all candidate SMBs, the next step was to select a subset of 2 to 8 SMBs that together satisfy the criteria described at the beginning of this section. There are two alternative approaches that can be used for this selection. In the first approach, a computer program was used to calculate the distributions of theoretical melting temperature for each SMB against each target sequence. From these melting temperature distributions, the computer program examines all possible combinations of SMBs (for example, a set of 6 SMBs) and scores the set for producing maximally dispersed melting temperature distributions for the entire group of target sequences. The set of SMBs that produces the best score for this parameter is then selected for use. In the second approach for selecting the SMB, the group of target sequences (such as those shown for Vβ regions in Figure 3) were first hierarchically clustered using a clustering program such as MEGA (Tamura et al. 2021 Molecular Biology and Evolution 38, 3022-3027). Hierarchical clustering analysis produces dendrograms that group similar sequences together and reveal the “evolutionary” relationships between the sequences. An example of such a dendrogram is shown in Figure 4. To identify groups of similar sequences, the dendrogram was sliced either vertically or horizontally. In Figure 4, four indicated groups were created by visual slicing. Within these groups, sequences are similar, but between them, they are quite different. Thereafter, an SMB was created, either for a randomly selected member of each group or for the consensus sequence from each group. The sets of SMBs that were selected by this procedure satisfy the criteria mentioned above. An example of four SMB probe sequences (SMB 1, SMB 2, SMB 3, and SMB 4) that were designed by this process to bind to portions of the TCR Vβ region are depicted in Figure 3, and their full SMB sequences are shown in Table 1. Docket No.: RU 2023-126 / FR: 096747.00494 When many similar sequences were aligned together, there were instances where a few sequences were identical. If all of the SMBs were to target the region where these sequences are the same, then discrimination between them would not be possible. Just such a situation occurs in the case of TCR β sequences TRBV6-1, TRBV6-2, TRBV6-4, TRBV6-5, and TRBV6-6 (Figure 3). In order to distinguish between these sequences, one or more of the SMBs in the chosen set can be moved to the left or to the right, until a region of these target sequences is found where they diverge from each other. Alternatively, an additional SMB can be used. Examples of such shifted target regions are shown in probes SMB 5 and SMB 6 in Figure 3. In conclusion, all of the SMBs in a set for one of the variable regions do not have to target the same sequence span. Experimental Determination of Melting Temperature Profiles of the SMBs Utilizing Synthetic Target Sequences After designing the SMB sets, they were synthesized with appropriate fluorophore and quencher pairs. Examples of fluorophore and quencher pairs, and their relative advantages, have been described in Marras et al. 2002 Nucleic Acids Research 30, e122. Other fluorophore and quencher combinations are known to persons skilled in the art. These SMBs can be purified and characterized as described earlier (Tyagi, et al. 2000 Nonradioactive Analysis of Biomolecules, second edition, C. Kessler, Ed., Springer Verlag, Berlin, Germany pp.606-616). Oligonucleotides corresponding to the portion of the gene that is the target of the designed SMBs were obtained from a custom oligonucleotide manufacturer. Each target oligonucleotide was then separately mixed with each SMB under salt and buffer conditions that were similar to the buffer conditions found in the PCR assay that was planned to be used. A slight molar excess of the target (2- to 10- fold) was utilized in these assays. The melting profiles (also referred to as thermal denaturation profiles) of the hybrids by the binding of each SMB to each target were then determined in a real-time PCR instrument that monitors the fluorescence of hybridization reactions while raising the temperature. To illustrate such an experiment, each of the 48 TCR β sequences listed in Figure 3 was mixed with each of the four SMBs (SMB 1, 2, 3, and 4; all having 2 nucleotide-long stems) whose sequences are shown in Table 1. The details of the procedure are presented in Example 4, and the resulting melting profiles are shown in Figure 5, left panels. Three control reactions, in which no target oligonucleotides were added to the SMB, were also included. Docket No.: RU 2023-126 / FR: 096747.00494 Figure 5 shows that the set of 48 TCR β sequences yielded highly diverse melting profiles. As to these melting profiles, consider, for example, the profile for targeting TCR β sequence TRBV13 that is marked in the Figure 5 plots. At low temperatures, the fluorescence intensity of the hybrids is high, because the molecular beacons are bound to their target (causing the stems to dissociate, preventing the interaction between the fluorophore and the quencher), resulting in the fluorescence of the fluorophore to be fully restored. For a certain range of temperatures, as the temperature is raised, the fluorescence of the TRBV13 SMB hybrids remains high, because these hybrids are stable in that range. However, at a specific higher temperature, their fluorescence begins to drop, because the molecular beacons are separated from their targets and their stems reform, enabling their fluorophore and their quencher to interact with each other, resulting in the suppression of their fluorescence. As the temperature is raised even further, the helical order of the SMB stems gives way to a random coil state. In this random coil state, the molecular beacons are more fluorescent than in their closed state, but significantly less fluorescent than in their target- bound state. The important temperature in these profiles is the melting temperature of the probe- target hybrid, which is the temperature at which half of the probe-target hybrids are dissociated and the other half of the hybrids are intact. This temperature can be identified by calculating the negative first derivative of the temperature vs. fluorescence intensity as a function of temperature (Figure 5, right panels), where the melting temperature appears as a peak. For example, the peak of TCR β sequence TRBV13 is labeled. In the control reactions, where no target was present, or in reactions containing targets that have insufficient complementarity to the sloppy molecular beacon probe, these peaks were not observed. Each target in the set generally yields a distinctive melting profile for each sloppy molecular beacon probe. To illustrate this point, the melting profiles of TCR β sequence TRBV13 with each of the four molecular beacon probes were labeled in Figure 5, left panels. Similarly, the peaks in the negative first derivative curves (melting temperatures) were also labeled (Figure 5, right panels). A comparison of the TRBV13 curves across panels demonstrates how the same target, when probed by different SMBs, yields very different melting profiles. One of the key aspects of the disclosed method is to use the combined set of SMBs in order to obtain a multiplex profile that characteristically identifies the target. Since multiple SMBs interrogate different aspects of the target sequence, their combined information is more accurate than that of each SMB by itself. Docket No.: RU 2023-126 / FR: 096747.00494 To illustrate the great diversity of melting temperatures that these sets of targets can yield with four different SMBs, the melting temperatures are shown for all targets and all molecular beacons on the same plot (Figure 6). This plot shows that almost all targets resulted in a unique set of melting temperatures that are distinguishable from each other. Notable exceptions were TRBV6-1, TRBV6-2, TRBV6-4, TRBV6-5, and TRBV6-6, which yielded a virtually identical set of melting temperatures with each of the four molecular beacons. However, this is because in the region where the four molecular beacons bind to these targets identical sequences occur (see Figure 3). Significantly, the specificity of the molecular beacon responses is illustrated when one compares the sequence of TRBV6-7 with these five sequences. It differs by just one nucleotide from the TRBV6-1 to TRBV6-6 set, and still, the melting temperatures yielded by the four molecular beacons for TRBV6-7 are different from those yielded by TRBV6-1 to TRBV6-6. This demonstrates that minute changes in the sequences result in detectable changes in the set of melting profiles. When a subset of the targets has the same sequence in a given region, such as in the case discussed above, the probe sequence of one of the molecular beacons can be changed so that it targets another region of that TCR sequence where the target sequences diverge from each other, or an additional SMB can be added. This can be accomplished by shifting the target region to the right by 9 nucleotides. This shift results in SMB 5, whose sequence is shown near the bottom in Figure 3. SMB 5 can either be used as a replacement for one of the four SMBs in the set, or it can be added to the set. Ensuring a dependable approach in designing SMBs for specific targets is crucial. Therefore, a correlation between the melting temperatures of the SMBs predicted by the DINAMelt computer program to those observed experimentally was assessed. The details are presented in Example 5. Figure 7 compares the melting temperatures determined experimentally with the melting temperatures predicted by the DINAMelt web server. The results demonstrated that there is a significant correlation between the two sets of melting temperatures, indicating that SMB sets can be designed bioinformatically without the need for extensive trials and testing. Using Multiple SMBs in the Same Reaction Tube to Create Compound Melting Profiles Another aspect of this invention is to use multiple SMBs for the same set of targets (exemplary sets are TCR α, β, V, or J) together in a multiplex reaction to create high-content Docket No.: RU 2023-126 / FR: 096747.00494 melting profiles. When multiple SMBs were used together, the set of SMBs yielded a compound melting profile for a given target in which the profiles of individual SMBs are superimposed on top of each other within the same curve. If the melting transitions of different SMBs for the same target are sufficiently separated from each other within the temperature range, they can clearly be resolved when seen in the negative first derivative plot. If, on the other hand, the melting transitions are close to each other, they may not be resolved. However, in both cases, a uniquely distinguishable profile is created for each target. An example of such compounding is illustrated in Figure 8. Gray curves at the top of Figure 8 represent four melting profiles for TRBV13 with SMBs 1, 2, 3, and 4 individually, and the black curve is the melting profile when the four SMBs were combined in one reaction. The negative first derivative plots of these profiles are presented at the bottom of Figure 8, which shows that in the compound profile, two of the four component melting transitions are clearly identifiable (from SMB 2 and SMB 4), whereas the melting transitions from the other two SMBs (1 and 3) appear to be compressed together. Nonetheless, the compound profile is information-rich, as it encompasses the response of each of the four probes that have their own separately distinctive discriminatory capacities. The details of how high-content compound melting profiles were created are described in Example 6. The diversity of the high-content compound profiles and their negative first derivative curves created by multiple SMBs used in a multiplex for all 48 targets and controls is shown in Figure 9. A visual examination suggests that these profiles are distinct from each other. However, to determine if each profile is sufficiently different from each other, and in order to ascertain whether the set of SMBs allows unambiguous identification of each different target from its compound profile even when statistical variations are present, a computer analysis needed to be performed. Such an analysis is presented below. Identification of Target Gene Segments from Melting Profiles A review of the compound profiles shown in Figure 9 indicates that a unique denaturation profile results for each target when four appropriately designed SMBs were used together in one reaction. However, it is important to computationally recognize and distinguish these profiles, because it is not possible for humans to distinguish thousands of similar-looking curves. Furthermore, humans would not be able to distinguish variations that arise due to minor sequence Docket No.: RU 2023-126 / FR: 096747.00494 variations, or that arise purely from probabilistic chance. In addition, it is difficult to obtain a single parameter from the compound profiles that would be indicative of the identity of the target sequence. Therefore, another key aspect of this invention is a system and method for developing machine learning models to identify target sequences from the melting profiles that are generated by the hybridization of sets of SMBs. The method involves obtaining multiple replicates of melting profiles from a set of multiple SMBs for one or more gene segments, as described above. Examples of such profiles are presented in Figure 9. These replicates were obtained by carrying out repeated experiments that obtained the melting profiles from the same set of targets and SMBs. These replicates introduced probabilistic variabilities that are expected in actual assays. The raw temperature vs. fluorescence curves recorded by the instrument were then processed to obtain the negative first derivatives, which are used for training machine learning algorithms. Alternatively, smoothed fluorescence vs. temperature curves can be utilized directly. Example 7 provides details of the data processing needed to implement machine learning- based identification of targets from profiles and discrimination between profiles of different targets. Briefly, the data from the negative first derivative curves, or from smoothed primary curves, was compiled into a spreadsheet, where each row represents the negative first derivative curves for a given target determined over the full range of temperatures. The last element in each row indicated the name of the target whose data appears in that row. The spreadsheet contained multiple rows for each target corresponding to the replicates. Multiple machine learning algorithms, such as the “Subspace Discriminator” in the “Ensemble Classifier” group, were trained by utilizing the data prepared in MATLAB. The predictor variables were defined as columns containing the melt data for each sample, and the response categories were defined as the target names. The trained model with the highest accuracy was then selected for further analysis. The accuracy of the selected model was demonstrated through a “confusion matrix” plot which visualizes the correct identification of the TCR β sequences. A confusion matrix plot for the data in Figure 9 is presented in Figure 10 using the Subspace Discriminator algorithm, which resulted in a discrimination accuracy of 96%. The dark gray quadrants in the plot represent correct identifications, while the light gray quadrants indicate misidentification events. The majority of Docket No.: RU 2023-126 / FR: 096747.00494 TCR β sequences were correctly identified along the diagonal, indicating high prediction accuracy. The numbers within the quadrants represent the frequency of correct calls, mostly being 6 (corresponding to the number of replicates used). Some quadrants, particularly for TCR β variants TRB6-1, TRB6-5, and TRB-6 (which possess identical sequences), show lower numbers. More details of this procedure are presented in Example 7. Overall, the results demonstrate the effectiveness and accuracy of machine learning in identifying TCR βV sequences. Similar analyses can be performed in other computational environments. For example, the Support Vector Machine Classifier in the scikit-learn open source data analysis library, which is the gold standard for machine learning in the Python computer environment, also resulted in a 96% accuracy of prediction utilizing the same data set. Partitions Multiple single T cells obtained from an individual can be distributed into small reaction vessels that are referred to here as “partitions,” where each partition generally receives a single cell. However, occasionally some partitions may contain more than one cell or may not contain a cell. T cells that are distributed into these partitions can be enriched from the donated blood or from other tissues via affinity purification. In the affinity purification procedure, antibodies against receptors on the T-cell surface, for example, CD3 and CD4, are tethered to magnetic beads and are utilized to selectively capture those cells. The affinity-purified cells are then purified further by flow cytometry before their distribution into partitions. When there is a need to analyze T cells from a solid tissue, such as a tumor, the cells in the tissue are first dissociated from each other using a mechanical or an enzymatic procedure, and then the T cells are isolated as described above. The partitions can be wells in a multi-well plate, aqueous droplets suspended in oil, or components of another suitable partitioning system known in the art. The partitions can be created by emulsification, microfluidics, or by other means known in the art. They may remain suspended in a three-dimensional space, spread on a two-dimensional surface, or passed through channels with a temperature gradient during the steps of amplification and / or detection (Figure 1). Each partition can further include cell lysis reagents, amplification reagents, detection reagents, and / or sloppy molecular beacon probes along with the cells from the beginning of the procedure. Docket No.: RU 2023-126 / FR: 096747.00494 Amplification of Recombined TCR Gene Portions from Single T cells in Partitions The next step after placement of the single T cells into partitions is the amplification of the recombined TCR genes or the amplification of the mRNAs encoded by the recombined TCR genes. This is accomplished by either PCR, recombinase polymerase amplification, transcription- mediated amplification, strand-displacement amplification, or another amplification technique known to persons skilled in the field. Compared to TCR genomic DNA amplification, TCR mRNA amplification is advantageous, because it is among the most highly expressed mRNA species in T cells, and these mRNAs do not possess introns. Amplification of these mRNAs is accomplished by including a high-temperature reverse transcriptase in the reaction, which creates a cDNA copy of each mRNA, which then serves as substrates for PCR. Target amplification, especially by PCR, usually includes an initial denaturation step, such as at 92 ºC, which naturally causes lysis of the cells, and renders the target nucleic acids accessible for amplification. The time devoted to this step can be increased to ensure efficient lysis. To help facilitate this process, mild detergents that do not negatively impact amplification can be included in the partitions. Primers that are suitable for amplification of the recombined TCR target genes or target mRNAs are included in the amplification mixture. These primer sets are designed such that portions of the recombined TCR genes, or mRNAs, are amplified for all T cells in their respective partitions, irrespective of which rearranged TCR sequences are present. Furthermore, these primers are designed such that all TCR arrangements are likely to be amplified equally well without bias. Persons skilled in the art will appreciate that the same primers will not amplify non- recombined TCR genes, because the distance between the two primer binding sites is so large that efficient amplification does not occur. It is advantageous to amplify a relatively small portion of the recombined TCR genes where the V and J regions meet. This region is often referred to as the complementarity determining region 3 (CDR3). This region is important for the antigen specificity of the TCR, and therefore holds high functional significance. To amplify this region, one of the primers is placed in the constant region that is close to the CDR3 region, and the second primer is placed in the variable region. The primer for the constant region is able to initiate DNA polymerization from all recombined arrangements, whereas the primer for the variable region initiated polymerization only Docket No.: RU 2023-126 / FR: 096747.00494 on particular TCR βV segments. Therefore, multiple V segment primers are needed at the same time to enable the amplification of all possible rearranged TCR genes. A suitable set of such primers has been described by Robins et al.2009 Blood 114, 4099-4107. Other suitable primer sets are discussed by Ch’ng et al. 2019 European Journal of Immunology 49, 1186-1199. The recombined genes of both the α and β chains are amplified simultaneously using a primer set specific for each. When multiple primers are used together, undesirable false amplicons are often produced, due to interactions between various primers. These false amplicons compete with the desired amplicons, and can therefore limit the sensitivity of the assays. Furthermore, biased amplification from some primer may also occur, which can negatively impact the accuracy of the quantification of the TCR repertoire. To suppress false amplifications and biased amplification events, several strategies can be utilized. In one of these strategies, the primer sets are designed so that minimal interactions are likely to occur between various primers, and then those primers that are the source of false or biased amplification are identified by deep sequencing of the amplification products. In a second strategy, an RNA residue is placed at a position several nucleotides inward from the 3’ end of the primer and the 3’ end is blocked. These modified primers are then used in the presence of a high-temperature RNase H that activates the primer by degrading the RNA residue when it is bound to its fully complementary target. False priming events fail to activate these primers. These and other types of primer optimization steps can be used for the amplification step of this invention. Alternatively, multiple V region primers can be replaced with fewer primers, or just one primer, designed to bind to the upstream constant region. However, this alternative is less desirable, because it will yield rather long amplicons that may be slower to amplify. In one example, this invention employs asymmetric amplification, such that the DNA amplicon strands that are complementary to the SMBs are produced in larger amounts than the corresponding complementary strands. This helps to ensure that the SMB targets are abundant and accessible for SMB binding, with little competition from the less abundant complementary amplicons. This asymmetry in PCR amplification was created by using a limited concentration of primers that are extended from the non-target amplicon strands, and an excess concentration of the primers that create the desired more abundant complementary strands by the time that PCR amplification ends. One method of asymmetric PCR is described in detail as “linear-after-the- exponential (LATE)-PCR by Sanchez et al.2004 Proceedings of the National Academy of Sciences Docket No.: RU 2023-126 / FR: 096747.00494 USA 101, 1933-1938. An additional benefit of asymmetric PCR is that the SMBs are not as readily degraded by an endonucleolytic TaqMan mechanism during the PCR, and they, therefore, remain intact for post-PCR melting analysis. Obtaining Post-PCR Melting Profiles from Each Single Cell Partition The disclosed method involves the determination of a melting profile in each partition after amplification. There are at least three different configurations of partitions that can be used to obtain melting profiles from the partitions. These configurations are referred to here as either “one-dimensional flow,” “two-dimensional stationary,” or “three-dimensional stationary.” The first two of these are illustrated in Figure 1. In the one-dimensional flow mode, which is suitable for aqueous droplets in oil suspension, the partitions are passed over a gradient of temperature as their fluorescence is monitored in multiple channels (Figure 1, right). Each droplet is tracked as it passes over zones of increasing temperature. This can be accomplished, for example, by the use of a linear array of light sources and detectors. The two-dimensional stationary mode is suitable for fixed partitions, such as wells and aqueous droplets suspended in oil spread over a surface. Here, the entire surface is heated while fluorescence was measured (Figure 1, left). This system is amenable to image-based recording of fluorescence changes as a function of temperature. The partitions (wells or droplets) can be computationally segmented (i.e., their boundaries are defined), and fluorescence intensity can be determined for each partition as a function of temperature. Examples of such systems are described by Athamanolap et al. (Athamanolap et al. 2019 Analytical Chemistry 91, 12784-12792 and by Luo et al.2022 Small Methods 6, e2200185). In the three-dimensional stationary mode, closely packed droplets in the bottom of a tube are used. Here confocal laser scanning, or light sheet microscopy, can be used to determine the fluorescence of each droplet even when it is located deep within the test tube. An example of such a system is described by Shum et al.2022 Analytical Chemistry 94, 17868-17876. The melting profiles are obtained simultaneously in several fluorescence channels. In the case of TCR α and TCR β amplification, four differently colored fluorescence channels are monitored, one each for TCR βV, TCR βJ, TCR αV, and TCR αJ. The reaction mixtures contain Docket No.: RU 2023-126 / FR: 096747.00494 SMB sets for each of these sets of targets. Each SMB set consists of multiple SMBs labeled with the same colored fluorophore, as discussed above, but the SMB sets for different targets have different and distinguishably colored fluorophores. As a result of this configuration, four melting profiles are obtained, one each for TCR Vβ, TCR Jβ, TCR Vα, and TCR Jα, respectively, for each partition. Targeting the Inter-primer Region or Targeting Complements of Tags on Primers During amplification, the specificity for the V and J segments can come from primers designed to be specific for individual V and J segments, or from SMBs that bind to individual V and J segments, or from combinations of the two. There are two methods in which the specificity gained from the primers can be combined with the specificity conferred by the SMBs. In the first method, the specificity generating 3’ ends of the V segment-specific primers are placed towards the 5’ end of the variable region (Figure 11 top) so that sufficient target variability remains in the amplified product. This ensures that the SMBs are able to bind to the variable region, and identify it through their melting profiles. In the second approach, the 3’ end of the V segment-specific primers are placed inwards within the variable region (Figure 11middle). However, in this second approach the 5’ tails of the primers contain tag sequences that “encode” the targets, thereby assigning unique codes to each target. The complements of the codes are arbitrary target sequences that can later be “decoded” by the binding and melting of the SMBs. After amplification, these tags are incorporated into the amplified product and then detected using a set of SMBs that bind to the complement of the encoding sequences (Figure 11 bottom). Even though these SMBs only recognize the complement of the encoded sequence, since there is a one-to-one relationship between the target and the code, the melting profile of the SMBs reveals the identity of the SMBs’ targets. Analysis of the Diversity of T-cell Clones and the Abundance of Specific T-cell Clones from Patients Considering the organization of TCR receptor genes depicted in Figure 2, one can calculate the number of combinations that can arise from a random assortment of different segments. Since the number of Vβ, Jβ, Vα, and Jα segments are 52, 13, 70 to 80, and 61 respectively, 2,886,520 to 3,298,880 combinations (52 x 13 x 70 (or 80) x 61) can arise. Note that this diversity is further increased by random nucleotide excisions and additions at the V, D, and J junctions during the Docket No.: RU 2023-126 / FR: 096747.00494 recombination process, though this source of diversity is not reported by the SMBs. Example 7 demonstrates how it is possible to identify essentially all Vβ segments using a set of SMBs labeled in one color. The rest of the segments Jβ, Vα, and Jα can similarly be identified using SMB sets that are specific to those segments and are labeled with fluorophores of three other distinct fluorescent colors. This results in four independent differently colored melting profiles that taken together report the identity of all four gene segments, Vβ, Jβ, Vα, and Jα. Since each profile explicitly identifies a particular target gene fragment, the four profiles together indicate which TCR arrangement is present in a cell. Since specific Vβ, Jβ, Vα, and Jα arrangements can be identified from 2.9 x 106to 3.3 x 106possible arrangements, the depth of this analysis is very high and covers all-natural diversity except the variability that arises from random nucleotide excisions and additions at the V, D, and J junctions during the recombination process. Since the diversity of T-cell clones is extremely high in blood, the vast majority of T-cell clones are found only once in a sample. Some clones may be expanded, and their expansion may vary from clone to clone. To quantify this diversity within a sample or population, a statistical parameter called Shannon’s entropy is often used, as described by Cui et al. 2018 Frontiers in Immunology 9, 2729. This parameter considers both the number of unique TCR arrangements present and their relative frequencies. The formula for calculating Shannon’s entropy is as follows: ^^^^ = −� ^^^^ × log2^^^^ where H is the Shannon entropy, frequency of each unique TCR arrangement in the repertoire. The summation (Σ) is taken over all unique TCR sequences. In this formula, ^^^^ × ^^^^ ^^^^ ^^^^2^^^^ represents the information content associated with each TCR sequence, and the negative sign ensures that the Shannon entropy value is non-negative. A higher Shannon entropy value indicates a more diverse TCR repertoire, with a larger number of unique TCR sequences, and a relatively even distribution of frequencies. Conversely, a lower Shannon entropy value indicates a less diverse repertoire, with fewer unique TCR sequences and / or an uneven distribution of frequencies. Shannon’s entropy can be used to compare the TCR repertoire diversity between different samples or populations, track changes in this diversity over time, and / or assess the impact of diseases or treatments on a patient’s TCR repertoire. Docket No.: RU 2023-126 / FR: 096747.00494 Shannon’s entropy is just one of many measures used in TCR profiling to characterize the diversity and complexity of TCR repertoires. Other diversity indices, such as Simpson’s diversity index or the Gini coefficient, may also be employed, depending on the specific research question or analysis approach. Identification of Public Clones When the TCR repertoires from many individuals with similar conditions, such as a disease state, are analyzed at sufficient depth, a few T-cell clones that are common between them can be identified. These clones are referred to as “public clones” (see de Greef et al.2020 eLife 9, e49900 ). Public clones are of high significance because they provide insights into the immune response to specific antigens or diseases. When multiple individuals share the same TCR sequence, it suggests that these individuals have encountered a common antigen, or have a similar immune response to a specific condition. The methods disclosed here can be used for the identification of these public clones. To perform such an analysis, T cells from multiple patients with a similar disease condition, such as having a confirmed diagnosis of an autoimmune disease, such as lupus erythematosus, or rheumatoid arthritis, are compared, and the clones that are common to each diagnosis are identified. Since only a few clones are closely associated with the disease state, they can be used as biomarkers for that disease. The disease condition can therefore be diagnosed in assays that focus on these marker clones. A simple way of creating such focused assays for the public clones is to use a subset of primers that are specific to the Vβ, Jβ, Vα, and Jα segments that are found in these clones in combination with a full complement of SMBs. This will result in amplification of only the public clones, and will not amplify the rest of the repertoire. Therefore, only those partitions that harbor one or more particular public clones will yield positive melting profiles. These focused assays can also be performed on public clones that have been identified previously by other methods such as next-generation sequence analysis. A key advantage of these focused assays is that since the rest of the repertoire is not amplified, the probability of finding and identifying the public clones is significantly increased. Example 2. Synthesis and Purification of Sloppy Molecular Beacon (SMB) Probes Docket No.: RU 2023-126 / FR: 096747.00494 The sequences of all of the SMBs used for these experiments are listed in Table 1 (SEQ ID NOs: 1 to 13). They were all labeled with fluorescein (FAM) at their 5’ terminus and with Blackhole Quencher 1 at their 3’ terminus. They were purchased from LGC Biosearch Technologies (Novato, CA) and purified by C-18 reverse phase high-pressure liquid chromatography in-house. The probes were resuspended in a 10 mM Tris-HCl, 0.1 mM EDTA, pH 8.0 solution at a concentration of 100 µM and stored at -20 ºC until further use. Target oligonucleotides (see Table 1, SEQ ID Nos: 14 to 61) were purchased from Integrated DNA Technologies (Coralville, IA), and were dissolved in RNase-free water at a concentration of 100 µM. The target oligonucleotide solutions were first diluted in water to a concentration of 10 µM and then further diluted in water to a working concentration of 1 µM. The stock and working target oligonucleotide solutions were stored at -20 ºC until they were used. Table 1. Representative sequences of SMBs and target oligonucleotides Sequence ( 5’ to 3’) SEQ ID No Docket No.: RU 2023-126 / FR: 096747.00494 PROBE AAGACAGCAGCATATATCTCTGCAGC 13 SMB 6 Docket No.: RU 2023-126 / FR: 096747.00494 TRBV23-1 gggAGACACGGCACTGTATCTCTGCGCCAgca 55 TRBV24-1 caaCCAGACAGCTCTTTACTTCTGTGCCAcca 56 For the SMBs, the underlined nucleotides represent mutually complementary 5 and 3 arm sequences that form the stem hybrids. For the target sequences, the three lowercase nucleotides at each end are present in the oligonucleotide target gene segments, but are not part of the sloppy molecular beacon probe binding region. Example 3. Optimization of the Stem Lengths of the Sloppy Molecular Beacon Probes Ideally, probe-target hybrid strength is the only determinant of the melting temperature of the hybrid formed by an SMB with its target oligonucleotide. However, it was observed that some of the targets that had many mismatches with the SMB did not register a melting temperature. However, this high “dropout rate” was rectified by decreasing the length of the stem in those molecular beacons. To verify this, four versions of SMB 3 were prepared. These versions differed from each other with respect to the length of their stems, which were either 6, 4, 2, or 0 base pairs long (Sequence IDs 1 to 4, Table 1). All of these SMBs possessed a 5’ fluorescein and a 3’ Black Hole Quencher 1. To determine the melting profile of each of these sloppy molecular beacon probes, mixtures (25 µl) of each SMB with each of the 48 targets (SEQ IDs 14 to 61, Table 1) were prepared. These reactions contained 100 nM SMB and 400 nM target oligonucleotide, and additionally contained 50 mM KCl, 10 mM Tris-HCl (pH 8.0), and 3 mM MgCl2. These are typical conditions under which PCR amplifications are performed. Three reaction tubes in which no target was present were also included as negative controls. Hybrid melting reactions were carried out using 0.2 ml white polypropylene PCR tubes (USA Scientific, Ocala, FL) in a Bio-Rad CFX96 Real-Time System spectrofluorometric thermal cycler. The thermal cycler was programmed to heat the reactions for 3 min at 95 ºC, after which the reactions were allowed to cool passively to room temperature. The thermal cycler was then programed to cool the reactions for 2 min at 4 ºC, and Docket No.: RU 2023-126 / FR: 096747.00494 to then increase the temperature in steps of 0.5 ºC to 95 ºC, holding each step for 10 sec while recording the fluorescence intensity. Fluorescence intensity as a function of temperature for each reaction is shown in Figure 12, left. Four plots are shown where the particular SMB that was present in those reaction is indicated. Each curve in these plots stems from one reaction and shows the response to one particular target. For each SMB, the plots are normalized such that at 85 ºC the fluorescence intensity is the same. The software of the thermal cycler automatically determines the negative derivative of fluorescence intensity with respect to temperature (-dI / dT) as a function of temperature. These instrument-generated plots are shown in Figure 12, right. From this data, the software automatically determines the melting temperature of the hybrid formed by binding of the SMB to the target oligonucleotide. The melting temperature data for all SMBs and all targets are presented in Figure 13. As described above, the fluorescence intensity in these melting profiles is generally high at low temperatures, because the molecular beacons are bound to their target oligonucleotides, which physically separates their fluorophore from their quencher, resulting in a strong fluorescent signal. As the temperature is raised further, the fluorescence intensity remains high, because the hybrids are stable. However, at a particular characteristic temperature for each different hybrid, fluorescence intensity begins to drop precipitously, which corresponds to the probe-target hybrid melting temperature, because the molecular beacons are separating from their target and their stems are reforming, enabling its quencher to interact with its fluorophore. As the temperature is raised even further, the helical order of the molecular beacon stems give way to a random coil state (which occurs at the stem melting temperature), somewhat increasing the background fluorescence intensity. However, the curves shown in the top left panel in Figure 12 (utilizing a molecular beacon possessing a 6-base pair stem) show that the fluorescence of many of the reactions remained low at lower temperatures. Prior to carrying out this experiment, it was expected that the probe-target hybrids would all form at these low temperatures, and their fluorescence intensity would be high. In addition, many other targets showed a trend of diminishing fluorescence intensity from a peak level at lower temperatures. Accordingly, when the negative first derivatives that correspond to Docket No.: RU 2023-126 / FR: 096747.00494 these curves were examined, it was apparent that very few peaks were observed below 50 ºC (see, for example, Figure 12 top right panel). These results show that when the sloppy molecular beacon probes possess a 6-base pair stem many targets failed to respond (i.e., they dropped out). These dropouts typically occurred when the probe-target hybrids possessed many mismatched base pairs. When the lengths of the stem in the sloppy molecular beacon probes were reduced to 4, 2, or 0 base pairs, the targets that were not responsive to 6-base pair SMBs began to yield clearly detectable melting transitions below 50 ºC (Figure 12, lower panels). The melting temperatures for all 48 probe-target hybrids with each SMB are shown in Figure 13. The melting temperatures for those probe-target hybrids that did not register a detectable melting temperature (missing in the negative first derivative plots) are shown on the X axis in these plots. These probe-target hybrids are “dropouts,” as the SMBs did not bind to these target oligonucleotides (or if they did bind to their target oligonucleotides, the flexibility of the hybrid that they formed containing many mismatches, combined with the greater length of their stem sequences, enabled their fluorophores to interact with their quenchers, significantly decreasing their fluorescence intensity). The number of dropouts was 21, 9, 4, and 3, for stem lengths of 6, 4, 2, and 0 respectively (Figure 13). This decrease indicates that smaller stem lengths in the SMBs yielded melting profiles from a larger number of targets. Overall, these results show that dropout rates decrease significantly when the length of the sloppy molecular beacon’s stem is decreased. A possible explanation for this observation is likely to be that when a probe-target hybrid contains many mismatches, it is not strong enough to compete with the stem hybrid in SMBs possessing long stems, but it is able to compete when the stem length is reduced. An alternative explanation is that when the probe-target hybrid possesses many mismatches, the hybrid becomes quite flexible, enabling the fluorophore and the quencher to interact; but when the stem length is reduced, the fluorophore and the quencher of the hybridized SMB are unlikely to be able to interact with each other. The particular sequence incorporated into the loop in the present set of SMBs is such that when the stem length is 0, the loop is still able to form a short stem-and-loop structure that provides a degree of quenching in unbound probes. Docket No.: RU 2023-126 / FR: 096747.00494 Example 4. Further Description of the Melting Profiles of Four Different SMBs Targeting TCR Vβ Regions The results shown in Example 3 demonstrate that each of the 48 TCR Vβ targets yields distinctive melting profiles with SMB 3, which possessed a two-base pair stem. However, there is a possibility that more than one target will sometimes yield a similar profile that will be difficult to distinguish. To minimize such possibilities, four SMBs that possess very different probe sequences were utilized. When these four sloppy molecular beacon probes were used together, they created compound melting profiles with a very high amount of information content. Following the guidelines that emerged from Example 3, four optimized SMBs, SMB 1, SMB 2, SMB 3, and SMB 4 (corresponding to SEQ ID NOs: 3, 5, 6, and 7), each possessing a two-base pair stem, were prepared. Their probe sequences were designed following the procedures described in the previous section entitled “Designing Sloppy Molecular Beacon Probes for T-cell Receptor Regions.” Melting profiles for each of these SMBs were determined for each of the 48 targets, while also including three replicate reactions that served as no-target controls. The reaction volumes, conditions, and the concentrations of the SMBs and target oligonucleotides were also the same as in Example 3. The results of these experiments are shown in the melting profiles in Figure 5. The plots in the left-hand panels in Figure 5 for each SMB are normalized such that at 85 ºC the fluorescence intensity is identical. The same results plotted in the right-hand panels in Figure 5 show the negative first derivative of fluorescence vs. temperature (-dI / dT) as a function of temperature. From this data, the thermal cycler software function was used to automatically determine the melting temperature of each SMB probe-target hybrid. Figure 6 shows the four melting temperatures (corresponding to four SMBs) for each of the 48 target oligonucleotides. Figure 6 shows that each of the 48 target oligonucleotides generally yields a unique and distinguishable set of melting temperatures. Example 5. Comparison of Predicted Melting Temperatures with the Experimentally Observed Melting Temperatures It is important that the procedure used to design the sequences of the sloppy molecular beacon probes for any given set of different oligonucleotide targets results in the design of probes whose predicted hybrid melting temperatures are quite likely to actually result in those predicted characteristic melting temperatures. Experiments were therefore carried out to determine how well Docket No.: RU 2023-126 / FR: 096747.00494 the melting temperatures of the SMBs predicted by the DINAMelt web server (Markham and Zuker 2005 Nucleic Acids Research 33, W577-581) correlated with the actual experimentally determined melting temperatures. The DINAMelt web server was used to calculate the melting temperatures of the DNA hybrids formed by the binding of each sloppy molecular beacon probe to its intended target oligonucleotide (parameters used: [oligo] = 0.1 µM; [Na+] = 60 mM; [Mg2+] = 3 mM; T = 25 ºC). These calculated melting temperatures for each of the four SMBs, SMB 1, SMB 2, SMB 3, and SMB 4 (corresponding to SEQ ID NOs: 3, 5, 6, and 7, respectively, each containing two-base pair stems) hybridized to each of the 48 target oligonucleotides are plotted against their experimentally determined melting temperature in Figure 7. The experimental melting temperatures used here were obtained from the results shown in Figure 6. Figure 7 shows that a very high level of correlation exists between the two sets of melting temperatures. This observation indicates that sloppy molecular beacon probe sets can indeed be designed bioinformatically, without extensive trials and testing. Example 6. Compound Melting Profiles for Single Targets with all four SMBs Present Together in the Same Reaction To demonstrate how multiple SMBs present in the same reaction and targeted against the same TCR Vβ sequences yield more detailed and discriminatory melting profiles than if the reactions were carried out with only one SMB, each of the four optimized SMBs (SMB 1-4, SEQ ID NOs: 3, 5, 6, and 7) were combined into the same hybridization reactions In order to illustrate how these results were interpreted, melting profiles are exemplified by one of the oligonucleotide target sequences, TRBV13 (SEQ ID NO: 47) in Figure 8. Gray curves in Figure 8 top represent four melting profiles for TRBV13 with SMBs 1-4 individually, and the black curve is the melting profile when the four SMBs were combined into one reaction. The negative first derivative plots of these profiles are presented in Figure 8 bottom, which shows that in the compound profile two of the four component melting transitions are clearly identifiable, whereas the melting transitions from the other two SMBs appear to be compressed together. Nonetheless, the compound profile is information-rich, as it encompasses the response of all four sloppy molecular beacon probes, each of which has a different discriminatory capacity. Docket No.: RU 2023-126 / FR: 096747.00494 Next, the compound melting profiles of the same set of SMBs with each of the 48 oligonucleotide targets, and with the three no-target controls, were determined, as described in Example 3. The reaction volumes, the reaction conditions, the method of data acquisition and analysis remained the same as in Example 3. However, the concentration of each SMB was 25 nM, and the concentration of the target oligonucleotide was 400 nM. Therefore, in these reactions, each SMB was present at 1 / 4ththe concentration that was used in Example 3 (where only one SMB was present in each reaction). The melting profiles and negative first derivative of the fluorescence vs temperature curves for all 48 reactions, each containing a different target oligonucleotide, are shown in Figure 9, as well as the results obtained with the non-target control reactions. Example 7. Developing Machine Learning Models for the Identification of Target Sequences from Melting Profiles The compound profiles shown in Figure 9 indicate that a unique denaturation profile was obtained for each target when four appropriately designed sloppy molecular beacon probes were used together in the same reaction. However, it is important to computationally recognize and distinguish these profiles, because it is not possible for humans to distinguish thousands of similar looking curves. Furthermore, humans would not be able to distinguish differences that arise due to sequence variations, or that arise purely from probabilistic sources. To develop a predictive machine learning model, multiple replicates of the compound melting profiles using a full set of SMBs for each of the TCR β targets were obtained. Six replicate reactions were carried out to obtain multiple independent melting profiles. From the raw temperature vs fluorescence curves recorded by the instrument, the negative first derivatives were obtained, and they were then used to develop a predictive model for identifying the target sequences that were present in the assay. The negative first derivative data was used to train the machine learning algorithm. Fluorescence vs. temperature curves can also be used. The data was compiled into a spreadsheet where each row represents the first derivative curves for the full range of temperatures for a given target (the black curve for TRBV13 in the bottom panel of Figure 8 is an example of this data). The last element of each row indicates the name of the target. In this spreadsheet there were 49 rows (48 for each target and one for the no-target control) for each replicate. Therefore, there were Docket No.: RU 2023-126 / FR: 096747.00494 49 x 6 rows for an experiment with six replicates. More replicates could also have been used for better predictability. Each row in this spreadsheet consisted of 184 columns, 183 for the values of the negative first derivatives, and the last one for the name of the target. Appropriate machine learning algorithms have been implemented in many computer languages. As an example, the MATLAB programming language was utilized. The Python computer environment can also provide equivalent results. Training data, prepared as described above, was imported into the MATLAB workspace. Columns 1-183 were defined as predictor variables, and column 184 (contained the names of the targets) was defined as response categories. MATLAB provided the possibility of training “classification learner models” using the same data set and then selecting the one with the highest level of accuracy. Accordingly, after training all available models, the one that provided the highest accuracy (96%) of prediction was selected. The “Subspace Discriminator” model in the “Ensemble Classifier” group was adopted. A MATLAB function was then created for this model that was able to predict the TCR β identity from each compound melting profile. The accuracy of the predictions made by this model was illustrated by a “confusion matrix” plot (Figure 10). In this plot, the dark gray quadrants signify the correct identification of each TCR β sequence, and the light gray quadrants represent misidentification events. The fact that almost all identified TCR β sequences are present along the diagonal indicates the high degree of accuracy of these predictions. The number in these quadrants indicates the frequency with which the correct call was made. This number is 6 (because six replicates were used) in almost all of the quadrants in the diagonal. It is important to note that there are several quadrants in the lower right that have numbers lower than six. These are for TRVB6-1, TRVB6-5, and TRVB6-6, which possess identical sequences, as was discussed earlier. Overall, the results in this example show that machine learning is a highly effective and accurate method for determining the identity of the TCR Vβ sequences that are present in a cell. Example 8. Obtaining melt curves from a large number of aqueous droplets in oil To demonstrate a method for simultaneously obtaining the melt curves of hybrids formed from Sloppy Molecular Beacon 1 (SMB) bound to oligonucleotide targets across numerous partitions, a system was developed for generating droplets from these targets. The targets, selected from Table 1, were mixed with SMB1 to create different kinds of droplets. These droplets were Docket No.: RU 2023-126 / FR: 096747.00494 combined, placed between two coverslips, and subjected to a controlled temperature increase while measuring the fluorescence of each droplet. The detailed procedure is described below. For this experiment, seven solutions were prepared containing 100 nM SMB1 dissolved in 20 µL of 1 x Bio-Rad ddPCR Supermix probe buffer (Bio-Rad, Hercules, CA). Oligodeoxyribonucleotides TRBV6-4, TRBV7-4, TRBV9, TRBV15, TRBV17, and TRBV19 were added to tubes 2 through 7 (400 nM final concentration), while tube 1 contained no target. The targets were chosen based on their expected melting temperatures with SMB1, roughly evenly distributed between 30°C and 65°C (Figure 5). Using a QX200 droplet generator with DG8 cartridges and droplet generation oil (Bio-Rad), each solution was partitioned into nanoliter droplets. In this system, the aqueous mixture of probes and targets passes through a microfluidic channel and merges with an oil stream, producing uniform and stable droplets (~0.9 nL) in an oil matrix. After creating the seven droplet suspensions, they were combined into a single mixture. Despite mixing, the droplets generally remained separate from each other. To monitor fluorescence while controlling the temperature, Vaheat coverslips (Interherence, Erlangen, Germany) capable of precise heating were used. The coverslips, coated with TiO2acting as a resistor, can rapidly reach and maintain specific temperatures. The coverslips have embedded sensors to provide feedback control of their temperature (visible in the Figure 14A image towards the upper right). The droplets were sandwiched between two coverslips by placing the droplet suspension on top of a Vaheat coverslip, and then placing a regular glass coverslip on top of the droplet suspension. This setup formed a monolayer of droplets due to surface tension. Moreover, to minimize droplet coalescence with the coverslips, a hydrophobic coating (Gel-Slick, Thermo Fisher Scientific, Waltham, MA) was applied to both coverslips. The droplet mixture sandwiched between the coverslips was then mounted on a Vaheat controller and placed on the stage of a Zeiss Axiovert M200 Microscope. Then the temperature was incremented from 35°C to 72°C in 4°C steps, pausing at each step for one minute. During each pause, an image was captured using a 4X objective in the fluorescein channel (which is the fluorophore that is attached to SMB1). Figure 14A shows an image of the droplets at 35°C, with the droplets varying in their intensities. Low-fluorescence droplets likely contained only SMB1, while the brighter droplets contained SMB1 plus target oligonucleotides. Larger droplets were Docket No.: RU 2023-126 / FR: 096747.00494 likely produced by the merger of more than one droplet, and these droplets were excluded from further analysis. In a series of images obtained at different temperatures, the droplets with a low initial fluorescence intensity increased their fluorescence intensity slightly (due to the opening of the molecular beacon stem). On the other hand, the fluorescence intensity of the brighter droplets decreased precipitously. This precipitous drop in fluorescence occurred at different temperatures in different droplets, indicating a target-specific behavior. To determine the melting profiles of the individual targets in single droplets, each droplet’s boundary was computationally delineated utilizing an in-house image-processing computer program designed to segment the single droplets in each image. The boundaries of the droplets were then adjusted (shrunk) to focus on the droplet cores, and the larger droplets were excluded because they likely arose from the merger of more than one droplet. The selected single droplets are indicated by polygons drawn on the image in Figure 14A and their labels are identified by numbers in this image. The mean fluorescence intensity at each temperature in the core of each selected droplet (within the polygons) was then noted, creating a table of fluorescence for each droplet as a function of temperature. The resulting fluorescence vs. temperature curves were then normalized to the same value at 72°C, because all of the hybrids can be assumed to be fully dissociated at this temperature, resulting in an identical fluorescence intensity (see Figure 5). The resulting 238 normalized melt curves were plotted in the left panel of Figure 14B. At higher temperatures, the fluorescence intensity of each curve decreases, which is indicative of the dissociation of the hybrid and subsequent decrease in the fluorescence of the molecular beacon. The resulting melt curves appear to be organized into several distinct clusters. For further analysis, a machine learning classifier (previously discussed in Example 7) was applied to identify DNA targets present in each single droplet. The classifier categorized the 238 melt profiles into 7 clusters (Figure 14B, right panel). In this figure, the melt profiles in each identified category are plotted using a unique line style selected from 7 available styles. The same 7 line styles were used in the left panel of Figure 14B, but there they were assigned randomly to different melt profiles. A comparison of the left and right panels of Figure 14B is informative, as Docket No.: RU 2023-126 / FR: 096747.00494 it demonstrates the classifier's accuracy in sorting the 238 melt profiles into 7 distinct categories, which correspond to the 7 types of droplets used to create the mixture. In this experiment 4°C, increments were used for simplicity. However, greater temperature resolution could be achieved by capturing images every degree or every half degree. While this experiment used a single SMB, higher content melt profiles could be obtained when several SMBs are combined together, as was done in Example 4. These high-resolution and high-content profiles will enable a distinction to be determined between many more targets. A person skilled in the art will understand that although this experiment utilizes synthetic TCR DNA sequences, targets can also be derived from amplified TCR mRNAs, or genes in single T cells. In such cases, the melt profile of each droplet will indicate the TCR arrangement of the encapsulated T cell, and the ensemble of profiles would represent the sample's TCR repertoire. In this study, 238 droplets were imaged over time in a single field of view. To increase the number of imaged droplets, a larger region encompassing more droplets can be scanned. This is routinely accomplished by moving the microscope stage in a programmed manner while capturing images. These images, or “tiles,” can then be stitched together into a mosaic that includes a greater number of droplets. Another strategy to increase the droplet count is to reduce the size of the droplets. Additionally, a third strategy involves imaging the droplets within a three-dimensional volume, such as a solution inside a test tube. Imaging droplet fluorescence in three dimensions within test tubes has been previously described utilizing light-sheet fluorescence imaging (Liao P et al. Three-dimensional digital PCR through light-sheet imaging of optically cleared emulsion. Proc Natl Acad Sci U S A (2020) 117:25628-25633.). This technique has the potential to rapidly observe up to half a million droplets while changing the temperature and thereby determine the melt profiles of DNA molecules encapsulated therein. The present invention is not to be limited in scope by the specific embodiments described herein. Indeed, various modifications of the invention in addition to those described herein will become apparent to those skilled in the art from the foregoing description and the accompanying figures. Such modifications are intended to fall within the scope of the appended claims.

Claims

Docket No.: RU 2023-126 / FR: 096747.00494 CLAIMS What is claimed is:

1. A method for profiling T-cell or B-cell receptor immune repertoires from multiple T cells or B cells possessing nucleic acid templates that encode chains of T-cell or B-cell receptors, comprising: providing a cell suspension comprising the cells suspended in a solution that comprises amplification and identification reagents, a plurality of primer pairs, a plurality of probes, and at least one polymerase for amplification and identification of a portion of receptor chains of the receptors; fractionating the cell suspension into a plurality of reaction vessels, such that each reaction vessel receives one cell along with the amplification and identification reagents; performing cell lysis and amplification; recording fluorescence signals from each reaction vessel, from each probe as a function of temperature, to generate melting profiles of hybrids formed by amplified targets and the probes; performing an analysis of the melting profiles obtained from each reaction vessel to identify an immune signature in sequences of the receptors that are present in each cell; and performing an aggregated analysis of immune signatures of the cells being analyzed, to determine an immune repertoire of the cells.

2. The method of claim 1, wherein the receptors comprise a T-cell receptor, and wherein the plurality of primer pairs comprise a first primer pair set that is capable of hybridizing to and amplifying a target region in the α chain encompassing both the Vα region and the Jα region, a second primer pair set that is capable of hybridizing to and amplifying a target region in the β chain encompassing both the Vβ region and the Jβ region; and these primer pair sets are chosen such that essentially all possible combinations of Vα-Jα in the α chain and essentially all possible combinations of Vβ-Jβ in the β chain are amplified.

3. The method of claim 2, wherein the receptors comprise a T-cell receptor, and wherein the plurality of probes comprise a first probe subset that is capable of hybridizing to a first amplicon in the Vα region and fluoresces in a first color; a second probe subset that is capable of hybridizingDocket No.: RU 2023-126 / FR: 096747.00494 to the first amplicon in the Jα region and fluoresces in a second color; a third probe subset that is capable of hybridizing to a second amplicon in the Vβ region and fluoresces in a third color; and a fourth probe subset that is capable of hybridizing to the second amplicon in the Jβ region and fluoresces in a fourth color; and the melting profiles of those hybrids are recorded and analyzed for each fluorescence color.

4. The method of claim 3, wherein a different number of probe subsets are used.

5. The method of any one of claims 2 to 4, wherein the cells are B cells and the receptors are B-cell receptors.

6. The method of any one of claims 1 to 5, wherein the immune signature of each cell is obtained through the use of one or more machine learning models.

7. The method of any one of claims 1 to 6, wherein the aggregated analysis of the immune signatures of all cells being analyzed is performed to obtain the immune repertoire of the cells.

8. The method of claim 7, comprising determining population characteristics of the immune repertoire comprising Shannon’s entropy, and identifying the expanded clones and public clones.

9. The method of any one of claims 1 to 8, wherein one or more characteristics of the immune repertoire are used for disease diagnosis or therapy management.

10. The method of any one of claims 1 to 5, wherein the probes comprise sloppy molecular beacon probes.

11. The method of any one of claims 1 to 5, wherein the length of each probe is between 10 to 50 nucleotides.

12. The method of any one of claims 1 to 5, wherein the plurality of probes comprise 2 to 25 probes.Docket No.: RU 2023-126 / FR: 096747.00494 13. The method of any one of claims 1 to 5, wherein the hybrids have a melting temperature of about 80 °C for probes that do not form mismatches with their target region and of about 25 ºC for probes that form the maximum number of mismatches with their target region.

14. The method of any one of claims 1 to 5, wherein the plurality of probes comprise probes having an internal stem or an external stem.

15. The method of any one of claims 1 to 5, wherein the amplification is performed by a polymerase chain reaction, a recombinase polymerase amplification, a transcription-mediated amplification, or a strand-displacement amplification.

16. The method of any one of claims 1 to 5, wherein the nucleic acid templates comprise mRNAs, and the amplification is performed by including a reverse transcriptase in reactions.

17. The method of any one of claims 1 to 5, where the amplification is asymmetric, producing a larger amount of the strand that is the target of the probes.

18. The method of claim 1, wherein the cell lysis is performed by heating prior to amplification.

19. The method of claim 1, wherein the plurality of reaction vessels are aqueous droplets.

20. The method of claim 1, wherein the step of fractionating the cells into the plurality of reaction vessels is performed by emulsification or microfluidics.

Images