I have an idea to build a simulated network based on the relationship between various factors, i.e., nodes, based on changes in probability, allowing variations, such as the prior probability and posterior probability of Bayesian networks, which can be extracted from the database in various proportions. www.biquge.infoHolistic medicine is the highest goal, and it is currently possible to diagnose, prevent, and treat common and critical diseases based on evidence-based medicine. The possibility construction of symptoms is dimensionality reduction, Taylor series decomposition, and the network structure formed by the collection of decomposition units of multiple symptoms corresponds to a certain disease. This is a system that can grow, run through the original experience in the early stage, and gradually match internally until a mature probability network is generated. Due to its probabilistic selective expression, which can generate new knowledge, the infinite possible combinations generated by the computer are left with valuable information after a certain amount of screening, which is a mechanism generated by high-quality combinations of natural selection. The coupling of information is a probabilistic selective expression, which is then reorganized into a new network structure. This is the thinking of the network.

Human beings are high-dimensional, that is, they have a high degree of selective expression, the accumulation of historical and cultural experience, and the indicators that are difficult to quantify, such as chest tightness and shortness of breath, are actually the coupling of multi-dimensional data, and the cognitive structure of human beings lies in the fact that they have a greater probability of choosing a certain optimal or suboptimal pathway. Machines also need to accumulate, so large-scale trial and error is inevitable

My goal is not to replace doctors, but to make them more efficient by reducing unnecessary repetition. At the same time, it also provides an opportunity to fill in the gaps through the computing power of the computer. Approaching the doctor's flash of inspiration with large-scale operations is also the eigenlevel of high-dimensional structures.

Diagnose, interpret, predict, make decisions, all in the network

Expert system, logical derivation is networked. The logical statement is not strictly ifthen mode, but ifxx%thenabc. At the beginning, there must be extremely large possibilities, and a large proportion of redundant data is required, so it needs to be gradually optimized by machine learning such as natural selection, and a fixed pattern can correspond to the real changes in the organism.

The possibility that known cases may have various effects can not be ignored and must be used as a reserve, the calculation of the path of the path of the specific mechanism of the specific mechanism of the specific symptoms of the specific disease is the only way for the computer to be superior to the doctor. No matter how small the probability is, what is possible will happen, and if this possibility is provided, the probability will be continuously increased according to the Bayesian formula in the specific logical reasoning chain, and the diagnosis can be made when a certain threshold is reached. Due to hierarchical coupling, sometimes we have to selectively decompose and recombine, i.e., find paths. Consider the heterozygosity of complex cases later.

Hierarchy, in traversal, dimension, in multiplying new variables. Points, multiple points

The possibility of construction of various behavioral representations, multidisciplinary i.e., hierarchical coupling: anatomically constructing the structure of the whole, the macroscopic level, the hierarchy below it, the system organs, tissues, cells are different orders of their Taylor series. The system is cyclic and tends to keep it intact and dynamically equilibrized. Organs, on the other hand, have their specificity and are regarded as expressions of multiple levels with varying degrees of coupling. The unidirectionality of tissues is stronger, and there is also coupling, and the specific mechanism is affected at the cellular level. The network built by the cells can be thought of as a collection of processes in the market that rely on the surrounding information

The explanation of the theory, observation is the first order that we generally recognize, we need to know that the probability network behind it is the most basic, the observation of the network should use a new scientific paradigm, consider the introduction of the yin and yang dichotomy of Bagua and its multi-order differentiation, the sequence 10100100101110101. Assumptions:1. The differentiation of times has a certain radius and boundary, and anything can be equated with a certain level or a set of specific levels (fixed point theorem and Fourier analysis)2. 10 is the basic division, like various periodic waves 3. A certain homology of the sequence is related to hierarchical coupling, and different degrees determine the distance of the distance4. Distribution is a fundamental function, and each level conforms to its probabilistic law at a large scale

Network generation and network existence are at the same level, which is a hierarchical coupling, macroscopically one is y, and the other is y'

Barabas's theory consists of two assumptions: first, that the number of nodes in the network increases from fewer to more, and second, that new nodes tend to be connected with more nodes with more existing connections. This network has the characteristics of "scale-free", that is, randomly removing some nodes and corresponding edges in the network will not change the topology of the network, so it is also called "scale-*******work" (the formation of high-dimensional structure is not affected by low-dimensional changes, unless quantitative changes cause qualitative changes, and also change the high-dimensional structure. In line with our cognitive structure, there is a certain degree of robustness)

It should be possible to hypothesize that the reconstitution of relationships, the generation and disappearance of which should be a dynamic process, not a change in the number of nodes but an overall change in the number of red positions

Propensity is important background, such as entropy increase, and there is a distribution that tends to be normally distributed, energetic. The Matthew effect of the number of node links, the complexity is one-way

Almost all biological networks are scale-free networks, which are probabilistic networks, and drastic changes in the whole will not affect the whole, and sometimes small changes in molecules lead to drastic changes in the network

Studies have shown that a 10% increase in the expression level of each molecule in the metabolic pathway leads to a 100% increase in final metabolite production, which is an overall change in the probability network of changes

From the perspective of the level of the network, there must be a certain level of immobility, and there should also be at the overall level of the crowd, but at this time it will lose its meaning. Because the high probability is a small range, there is also a certain distribution here. Cancer-associated mutations are prominently present in genes for specific signaling pathways. That is, there is no commonality in a single mutation, a single gene, but there is a strong tendency at the network level. The target does not have to be a single gene, it can be a specific pathway or network, which is a second-order structure and is more robust

Marginal algorithm for markers

Using the method of gene chip detection, the specific expression was first discovered. Next, by detecting the expression of lncRNAs during fasting and resumption of diet in mice, one of the three lncRNAs was screened to have significantly decreased expression after fasting, and the expression was also restored after resuming diet, that is, lncRNA, lncstr, which was related to the ability level in mice. At present, there are many studies on lncRNA discovery, but there is little work to analyze the function of lncRNA, and many of the work cannot be deepened only after making an expression profile of lncRNA or finding differentially expressed lncRNA. An important reason for this phenomenon is that researchers are not familiar enough with functional examination studies to find suitable conditions to screen for certain functionally relevant LNCRNAs. Therefore, this work has given bioinformatics researchers a good inspiration, to pay more attention to biological problems and regulatory mechanisms, and to communicate or cooperate with biologists, which is conducive to the in-depth research work.

Through the comprehensive analysis of microarray data under different conditions in the NCBIGEO database, Zhao Yi's research group found that the genes involved in multiple metabolic regulatory pathways were related to the expression of LNClstr (with great correlation), and further found that the expression of an important rate-limiting enzyme (CYP8B1) in bile synthesis was positively correlated with the expression of LNCLSTR, which is likely to be the target gene downstream of LNCLSTR. This hypothesis was confirmed by subsequent experiments, and the molecular mechanism of LNCLSTR preventing TDP-43 from binding to the CYP8B1 promoter by binding to TDP-43 was completed, thereby relieving the transcriptional repression of CYP8B1 by TDP-43. Unfortunately, the homologous sequence of LNCstr has not been found in human cells, and it is not known whether there is the same LNCRNA in human cells, or whether there is LNCR with different sequences but performs the same function.

The data are integrated with a level of reference for complementarity, which is quantified into a certain sequence (Fourier analysis) and then considers its homology and specificity, which emphasizes high-dimensional structures, such as Mendel's segregation rate and free combination rate of gene quanta.

Specific combinations have a higher probability of forming specialized regions, such as CPG islands, N6-methylated regions, and the like, which is the second-order level

The partial derivative is the ratio of a whole to a part, i.e., a variable, and provided that it is a continuous space, then its order of higher orders is equivalent. The integral is path-independent and is a high-dimensional relationship that is coupled into a loop. The relationship between surface division and reintegration of the curvature, the sum of partial derivatives of multiple dimensions

Variations in dimensions, such as Green's formula, can be used to calculate low-dimensional calculations where the dimension is equal to the higher dimension

The theory of big data is highly abstract, such as the physical force, the relative action, relying on the emergence of probability, what are the laws of mathematics?

;