The hierarchical coupling of the network is not random, but a meaningful pattern formed after a certain mechanism is filtered, like the collapse of the wave function. Because randomness inevitably produces a certain distribution, and the specific segments of the distribution function have stronger environmental adaptability than other segments, it is like a leftover from the process of natural selection, which is evolution. The collapse of this path is also a prerequisite for further development, that is, the forward probability of the Bayesian formula, which explains the existence of path dependence to a certain extent.

The specific mathematical model is sequence matching recognition, such as the recognition of ligands and receptors between cells. This is the recognition method of the reference immune system: 1 level of recognition, identification of cells, tissues and organs, 2 levels of differentiation, similarity of the concepts of central immune organs and peripheral immune organs 3 levels of complementarity, which is essentially a multi-level competitive game to form a balance, innate immunity and adaptive immunity, 4 levels of selective expression, immune defense, immune self-stability, immune surveillance

Sequence recognition is multi-layered, not just about structure, but also about structure-based functionality. Let's start with the basic structure, and then selectively express it as the identification of function in the traversal of the layers formed on this basis. According to the assumptions of the fractal structure, it has a certain similarity. Like logic circuits, different levels of packaging require certain interfaces.

Antigens and antibodies. The antigen binds to the TCR or BCR of T cells and B cells, promotes their proliferation and differentiation, produces antibodies or sensitized lymphocytes, and binds to them, and then exerts immune effects. This is the path formation of the network. The specific pathway formation is the various mechanisms we have explored, such as the thymus-dependent antigen (TD-Ag) that relies on T cells when stimulating B cells to produce antibodies, and the thymus-independent antigen (TI-Ag) directly stimulates B cells without T cell assistance, can produce IgM antibodies, and has no immune memory. This is the hierarchical coupling of the network.

So we come to the cellular level, where a specific combination of T cells, B cells, macrophages, and so on can have certain effects, that is, various immune mechanisms.

When it rises to the level of tissues and organs, the thymus gland is closely related to cellular immunity and humoral immunity, and the spleen is the largest peripheral immune organ in the human body.

After a bottom-to-top experience, it's time to start a new top-to-bottom model.

In a certain sense, we can think that cells can selectively express the sequence of their surface marker molecules, such as T lymphocytes with T cell differentiation antigens: CD3, CD4, CD8, CD28. Different cells can be defined according to this definition for their unique identity, just as different points in space are defined. Then it doesn't have to be too cellular interactions to abstract into interactions of these sequences. When we collect enough data, we can find a certain pattern, that is, the molecular mechanism of immunity, which is the law of statistical level.

TCRs do not directly recognize epitopes, but can only specifically recognize antigenic peptide-MHC molecular complexes on the surface of antigen-presenting cells or target cells. This is also a high-dimensional recognition mode, which is an interface with special requirements and contributes to the overall modularity.

T cells are divided into two types: CD4+ and CD8+, CD8+ T cells are effector T cells that mediate cellular immunity, and can specifically kill target cells carrying sensitized antigens after being sensitized by antigens. Th1 cells and Th2 cells differentiated from CD4+ T cells, Th1 cells secrete IL-2 and IFN-γ, and Th2 cells secrete IL-4, 5, 6, and 10. The differentiation of cells (multi-level, T/B cells are also a type of differentiation of immune cells), and the traversal of its levels can form different modes, such as cellular immunity and humoral immunity, and so on. B cells can not only exert specific humoral immunity functions through the production of antibodies, but also are important antigen-presenting cells. This creates a certain complementarity of concepts. We believe that this is the source of the compensatory mechanism of the network, that is, the hierarchy can build a certain equivalence relationship.

The preference of molecular distribution, such as the distribution of MCHII-like molecules on the surface of dendritic cells, is also an interface that contributes to the orderly conduct of immune function. We can abstract the functionality of a particular sequence.

Antibodies are part of sequence recognition and exert humoral immunity by binding specifically to the corresponding antigen (sequence recognition at the structural level).

The binding of antigens and antibodies is a sequence recognition at the structural level and then leads to changes in other functions, while cytokines exert their functions through the recognition of similar functional levels, such as interleukin-1, 6 (IL-1, 2) can cause fever is a macroscopic level relationship, which can be further decomposed into a detailed relationship sequence

Leukocyte differentiation antigens and adhesion molecules, etc., are other levels of recognition, and different recognition antigens produce different effects.

The above levels of ergodic coupling can construct a high-dimensional structure, i.e., the immune response. The specific path of the selective combination of different cell molecules is the macroscopic three stages of recognition, activation and effect. The mechanism of its selective expression can be idealized into a certain sequence, such as its selective expression can form an exemption

Plague tolerance and hypersensitivity. And various immune diseases are a way for us to study the immune system, that is, we look at the effects that the loss of each part may cause to speculate its role in the network, such as immunodeficiency.

Selective expression of the immune network can be tumorigenesis, i.e., selective expression (inhibition/activation) at different links, which can be expressed as a certain sequence. It is instructive for our specific treatment, i.e., to choose a meaningful combination to exert influence so that the overall balance moves to a healthy state. Theoretically, drugs should anchor the immobility of these sequences, such as anti-CD20 monoclonal antibodies, to treat non-Hodgkin lymphoma. Of course, in general, we still need to find multiple specific targets to work together, such as cocktail therapy.

Hardy-weinberg's equilibrium law based on stochastic assumptions reveals that the frequencies of each gene and each genotype remain unchanged for generations, which is an ideal of the real world, which can be used as a boundary to make a more quantitative explanation of the real world. We move from the assumption of random independence to the hierarchical coupling.

Genetic algorithms are the process of mimicking natural selection, allowing more adaptive sequences to multiply on a large scale, repeating this process to see a certain evolution, just like our real world. Because of the variation, Hardy-weinberg's equilibrium law is ideal.

Genetic hybridization is a macroscopic sequence operation, which can be idealized as a redistribution of coarse-grained genes, which is also a fixed point of thought. Of course, this is the operation of a single element of the sequence, and we aim for a whole line of the sequence of macro factors, which requires the judgment of indicators and so on, and we intend to refer to the scoring matrix idea of bioinformatics.

The disease itself can be expressed as a coupling of multiple sequences, and we can diagnose, prevent, and treat by processing the sequences. Of course, the premise of this is that we build this database, just like the genome measures the genetic sequence of a cell. Ask yourself, we have come to the same path as functional genomics, but the sequences we choose to operate are very inconsistent, they are from the bottom to the top, hoping that the study of genes has been traversing to form a high-dimensional study of human body functions, and what we hope is to continuously decompose the complex changes of the human body into different ranges of levels from the top to the bottom. Of course, we will meet eventually.

Compared to other diseases, we believe that genetic diseases are more subdued, i.e., the concept of compensatory in the body is not very useful, allowing us to observe the role that specific changes can play at the macro level. This high probability correlation is more friendly to mathematical derivation, such as Down syndrome is trisomy 21. Let's not consider the interaction between environmental and genetic factors for a moment. Of course, we only hope to be able to construct a sequence mathematical structure based on this to account for this interaction, but there is nothing we can do about it at the moment.

The classification of blood groups is an example of the sequences of the organism's network, and the competition of these sequences can be explained by the development of genes and genotypes, the Hardy-weinberg law of equilibrium.

Just as thalassemia, sickle cell anemia, etc., are relatively low-level and close to the central law, our application of the central law is of great significance for the solution of the disease. This is the idea of gene therapy. We don't have to do gene transduction, but we can also play a certain role in the regulation of gene expression through RNA. For example, we can continue to exert influence by placing a bacterial factory inside and outside the body, and we may be able to maintain a healthy lifestyle. So we need to find the object of the computation, which has always been to find the exact genes and enzymes, and we think that this is a fixed point, which is really important, but we can take a certain alternative way, which is to find a larger range of related genes and enzymes to accumulate their influence to a certain height through the Bayesian formula.

So what if this sequence is understood by observation? We need to refer to some relations of the central law to construct this mathematical structure. That is, we need certain reproducible studies to prove its equivalence with general structural sequences (ACGTs). Can the complexity of the algorithm be used as a measure, because the coupling and pattern emergence within the sequence are also important.

Genealogical analysis can be seen as the analysis of fixed points, which can be treated as part of a sequence, and of course the study of the sequence as a whole is large-scale and massive. The dominant recessiveness of a single gene is the result of the selective expression of the sequence. Complete dominance and incomplete dominance and co-dominant and irregular dominance and delayed dominance and subordinate dominance are all such selective expressions, that is, they are all expressions of probability.

Polygenic genetic disorders are a good example of sequence operations. It can be broken down into sub-layers of modules: each pair of alleles plays a role in the formation of inherited traits, co-dominance, and can add up to form a significant phenotypic effect (as in calculus). The various distributions are a function of sequence differentiation, and setting a certain threshold according to their specific location can be reduced to a single gene.

The specific position determination is the processing of meaningful sequences, and various fluorescent staining and other means can be determined macroscopically. Of course, SNP can be seen as the search for seed sequences of the BLAST algorithm.

Due to the hierarchical similarity of the fractal structure, we are able to discover various regularities at the individual level at the frequency of the group level. The law of equilibrium should be universal, which can accept the equilibrium achieved by the competitive game of the sequence. Moreover, the probes used in the definition of genes are also coupled by sequence similarity, but the diagnosis of genetic diseases requires higher-dimensional sequence matching, such as a series of symptoms and detections.