JP2020062015A - Living organism information processing device and method as well as program - Google Patents

Abstract

To predict changes in the state of living organisms and to enable control.SOLUTION: An expression level of molecules in a living organism is measured at specific time intervals, the measured time series data is divided into periodic components, response to environmental stimuli components, and base line components. Constant regions of the time series data is identified from fluctuations in the base line components, the amplitude of periodic components or fluctuations in the period, and a causal relationship between the identified constant regions is identified. A relationship between fluctuations in the internal environment and the external environment is identified, and from the causal relationship between the identified constant regions, changes in the state of the living organisms are inferred. Furthermore, from the relationship between fluctuations in the internal environment and the external environment, the onset of a disease in the living organism can be inferred. The molecules can be molecules in the blood or molecules in the culture medium.SELECTED DRAWING: Figure 2

Description

  The present technology relates to a biological information processing apparatus and method, and a program, and particularly relates to a biological information processing apparatus and method, and a program that predict and control future changes in human biological conditions.

  The loss of productivity in the development of new therapies, diagnostics and preventions has become a major problem in solving social issues related to health. This decrease in productivity is due to the fact that the vast amount of knowledge obtained in basic biology cannot be effectively used to solve clinical problems (Non-Patent Document 1). According to the latest international survey, the world's current diabetic population has increased to 285 million. By 2030, this diabetic population is projected to exceed the population of North America, more than 435 million. This means that the worldwide prevalence of diabetes in adults is approaching 7%.

Humans consist of 60 trillion cells, and each cell has the complexity of storing 6 billion base pairs of DNA (deoxyribonucleic acid) information, and it is said that it will undergo 10 ^{^} 16 cell divisions in its lifetime. It is a dynamic system. In addition to millions of base substitutions, a diversity of DNA structures such as deletion, duplication and inversion of DNA sequences is observed between two independent individuals.

  Hierarchy is one of the characteristics of biology. The human body is based on "cells" composed of molecules such as DNA and proteins, and is further characterized by subsystems, tissues, and organs composed of various cells. This hierarchy diversifies the time scale of the response to environmental stimuli. Intracellular information transmission proceeds in milliseconds or seconds, information transmission between cells proceeds on a minute or time scale, and cell proliferation and differentiation processes proceed on a date / time basis. Such a variety of time scales causes a time delay in the living body.

  In such complex and multifaceted life phenomena, basic biology has clarified the causal relationship between the molecules that compose the living body for the purpose of representing the living thing as a mechanical system. This has been performed by a method of selecting and changing only one specific molecular parameter from a large number of parameters.

  On the other hand, clinical sciences have focused on treating the disease by alleviating the associated symptoms.

  In addition, in order to see clinical effects, it has been practiced to infer empirical rules such as rules and relational expressions between interventions such as medication and clinical outcomes from biological indicators at specific times. .

  An important issue in modern medicine is to provide appropriate health care and treatment for each person. Even treatments and medicines that have been statistically proven to be safe and effective will be observed in some patients with side effects and ineffectiveness once they are launched and used in a large population. Become. This is due in part to the genetic diversity of treated patients. Regarding the difference in disease incidence due to genetic diversity, analysis has recently been advanced by "Genome wide association study (GWAS)" (Non-Patent Document 2).

Nature 453; 840-9, 2008

J. Hum. Genet. Doi: 10.1038 / jhg.2010.19

  However, in the current clinical practice, in order to measure a limited biological index only for a limited time, and to statistically average it, it is possible to obtain various biological indicators that various people who make up the population acquire over time. The characteristics are eventually discarded. In addition, while fundamental biology focuses on microscale issues, clinical studies show that individual-level macroscale issues such as physical disability, loss of physiological function, certain diseases and physical conditions. Is treated as a clinical outcome. There are gaps between strata in basic biology and clinical research.

  In addition, current treatment strategies have to be constructed by ignoring the diversification of diseases acquired over time, and thus both treatment strategies are symptomatic treatments. However, the disease does not return to its original state of health by alleviating the symptoms.

  These problems mean that current basic biology and clinical concepts cannot predict and control future changes in human biological conditions, making it difficult to develop effective preventive and therapeutic methods. is doing.

  The present technology has been made in view of such a situation, and makes it possible to predict and control future changes in the human biological state.

  A biometric information processing method or program according to an aspect of the present technology includes a lifestyle information acquisition step of acquiring information on a lifestyle of a target person, and information on a history of genetic modification or epigenetics modification of the target person. Acquiring history information acquisition step, acquired information about the lifestyle, and information about the history of the genetics modification or epigenetics modification, a similarity analysis step of analyzing the similarity with other than the subject. Including. A biometrics information processing device of one side of this art is a biometrics information processing device corresponding to the biometrics information processing method of one side of this art.

  In one aspect of the present technology, information regarding the lifestyle of the subject is acquired, information regarding the history of genetic modification or epigenetics modification of the living body of the subject is acquired, and information regarding the acquired lifestyle is acquired. And, the information regarding the history of the genetics modification or the epigenetics modification is analyzed for the similarity to those other than the subject.

  As described above, according to one aspect of the present technology, it is possible to predict and control future changes in the human biological state.

It is a block diagram showing composition of one embodiment of a living body information processor. It is a flow chart explaining health condition prediction control processing. It is a figure which shows the model regarding a causal relationship which makes a stationary region a node. It is a block diagram which shows the structure of other embodiment of a biometric information processing apparatus. It is a flowchart explaining the prediction control process of the cell state of a cultured cell. It is a figure which shows the structure of a cell model. It is a figure which shows the state change of a cell. It is a figure explaining the causal relation of information. It is a block diagram which shows the structure of other embodiment of a biometric information processing apparatus. It is a flow chart explaining other health state prediction control processing. It is a figure which shows the model regarding the causal relationship of lifestyle. It is a figure which shows the matrix showing the history of cell memory.

Hereinafter, modes for carrying out the present technology (hereinafter, referred to as embodiments) will be described. The description will be given in the following order.
1. Progress 2. Definition 3. First embodiment 4. Second embodiment 5. Third embodiment 6. Fourth embodiment 7. Fifth embodiment 8. Sixth Embodiment 9. Seventh embodiment 10. Eighth embodiment 11. Ninth embodiment

<Progress>

  First, the process of the present technology will be described as follows.

  The developers of the present technology have eagerly studied a method for representing the time evolution of the human body and predicting and controlling the future direction of change. As a result, in addition to the three parameters of genotype, phenotype, and external environment that have been conventionally used to formulate life, a new parameter called “time-developing cell memory” is introduced, It has been found that the problem of the living human body can be solved. In addition, in order to formulate the "time-developing cell memory", the developers of this technology are able to identify epigenetic modification and genetic modification of the cell memory of transcription factors that have the characteristics of the bistable switch expressed in each cell. We paid attention to the fact that it is changed with time due to modification. Epigenetic modifications consist of DNA methylation and histone modifications, and genetics modifications consist of DNA mutations and structural changes.

  The developers of this technology say that genetic modification and epigenetics modification change the cell state by modifying the expression level of the target gene product controlled by the bistable switch. Was built. Furthermore, the developers of the present technology have constructed the concept of "chromosomal state" consisting of transcription factors that work on chromosomes, genetics modification, and epigenetic modification according to "a time-evolving cell memory model".

  Next, the developers of the present technology, in order to estimate this “chromosomal state” from the observed values of the whole body state or the local state, are three different hierarchies constituting the human biological system: the whole body state, the local state, and the “ "Chromosomal status" was linked. For this reason, the developers of the present technology have constructed a macro model, a meso model, and a micro model corresponding to each hierarchy of the general condition, the local condition, and the “chromosomal condition”. Furthermore, the developer of the present technology proposes a method of connecting and formulating a macro model, a meso model, and a micro model on a cell-by-cell basis by using this “biological hierarchy connection model”.

  Next, the developer of this technology considers the development of a method that integrates the "biological hierarchy connection model" into a time series model in order to observe the "chromosomal state" that is responsible for the time evolution of the human living body, and We found a new integration method using "model". Then, the developer of this technology uses the "biological state space model" to divide the time series data of the molecular marker reflecting the general state into a periodic component, an environmental stimulus response component, and a baseline component. , And proposes a method for extracting the “chromosomal state”.

  Next, the developers of the present technology have earnestly studied the development of a control model using the data of the time evolution change of the human living body, which can be represented for the first time by the present technology. And the developers of this technology found the concept of "constructed change", which means that the input from the first external environment changes the responsiveness to the input from the subsequent external environment, and "the time-developed cell memory We propose a "dynamic construction model" in which "is controlled by" constructed changes ".

  In addition, the developers of this technology have conducted extensive studies to apply the "dynamic construction model" to clinical problems, and have calculated the time-series data regarding the expression level of molecules produced in the living body as "the biological state space model". We proposed a method of "Biological Local Stationary Region Model" for identifying stationary regions using. Furthermore, the developer of the present technology proposes a method for tracking a biological state and analyzing a causal structure by using, as a node, a biological local constant region found in a time-series change in the expression amount of a molecule produced in the living body.

<Definition>

  Next, the definition of terms used to describe the embodiments of the present technology will be described.

  The time evolution of the human body refers to the process by which the state and function of the human body change irreversibly with time. Humans die from the moment of fertilization through irreversible changes such as development, birth, growth, and aging. The onset of disease also develops into the pre-stage of onset, the onset, specific physiological dysfunction, specific physiological loss, physical disability, and death, which have undergone potential changes. This means that the time between one instant and another in the human organism is not homogeneous.

  "Time-evolving cell memory" indicates that the cell memory constructed by transcription factors is altered by genetic and epigenetic modifications introduced as a result of environmental input to cells or by chance. Conventionally, cell memory has been defined as a phenomenon in which a change in the state of a cell acquired by an external stimulus is maintained even after the external stimulus disappears (Cell 140: 13-18, 2010). The cell state is represented by the type, amount and modification characteristics of the molecule expressed in the cell, and the maintenance and change of the cell state are governed by transcription factors.

  In other words, conventional cell memory gives the concept that the cell state gives a specific equilibrium state due to the characteristics of the bistable switch of the transcription factor, but it can only give the switch on and off for a uniform time. It was However, in reality, genetic modifications and epigenetic modifications that are introduced over time change the cellular state provided by transcription factors over time, making the cell state much more than the cell state assumed by transcription factors alone. Diversify.

  In other words, in "time-developing cell memory," it is possible to formulate the diversity of cell states by the inheritable history of genetics modification and epigenetic modification, and "time-developing cell memory" It is essentially different from conventional cell memory in that it can handle heterogeneity.

  Transcription factors form the molecular basis of conventional cell memory. When the transcription factor circuit exhibits a non-linear and bistable characteristic, the molecular state is in the on or off equilibrium state, and its memory is inherited even after cell division. The affinity formed between the binding sites of transcription factors and promoters, the coordination between multiple transcription factors or multimerization imparts non-linear characteristics to the transcription factors.

  This non-linearity imparts a threshold-like characteristic to the reaction of the transcription factor to resist transient disturbance and maintain the expression level. In addition to this, transcription factors that acquired a sufficiently large Hill coefficient by positive feedback or positive or negative double feedback acquired the characteristics of a bistable switch, and the state change locked even after the original input disappeared. To be done. That is, what constitutes the skeleton of conventional cell memory is the type, amount, and post-translational modification of the transcription factor expressed in the cell.

  The molecular basis of "time-evolving cell memory" is genetic modification such as single nucleotide polymorphism observed in the human genome, partial deletion of DNA sequence, duplication, and copy number change, and DNA methylation. , Histone modification, protein denaturation and other epigenetic modifications (Nature Review Genetics 7: 85-97, 2006, Cell 128: 655-658, 2007).

  The genetics modification and the epigenetic modification modify the conventional function of cell memory by directly or indirectly changing the parameters related to the non-linearity of the transcription factor and the characteristics of the bistable switch.

  Regarding genetic modification, the relationship between single nucleotide polymorphisms and disease incidence has been analyzed by whole genome linkage analysis. Using that analysis method, 19 genes have been identified as related genes of type 2 diabetes. However, only 1% of diabetic patients carry this mutation (Nature 462: 307-314, 2009). Similarly, only 3% of BRCA1 / 2 mutations are observed in all breast cancer patients. Studies comparing the lifespans of monozygotic and dizygotic twins have shown that the contribution of gene sequences to lifespan is about 15% to 25% from epidemiological studies (Hum. Genet 97: 319-323, 1996).

  Epigenetic modifications are defined as inheritable alterations in gene function that are not accompanied by DNA sequence changes. DNA methylation and chemical modification of chromatin proteins are the molecular entities of epigenetics. In addition, modification of cell function by denaturing proteins such as prion protein and amyloid protein is also classified as epigenetic in a broad sense. The concept of epigenetics was proposed by Waddington to explain how genetic information is phenotypically transformed in the process of development.

  However, the function of epigenetic is not limited to the establishment and maintenance of tissue- and cell-specific gene expression, but nutritional status, social stress, and stimulation by chemical substances introduce epigenetic modification. Human epidemiological studies have reported that environmental irritation received during fetal and neonatal influences the incidence of post-adult chronic disease through epigenetics (Stem Cell Res. 4; 157-164, 2010).

  Such epigenetic modifications are called environmental epigenetics to distinguish them from developmental epigenetics associated with cell differentiation. The developer of this technology has found that the function of environmental epigenetics is regarded as "the function of updating DNA information based on changes in external environmental information".

  Genetic and epigenetic modifications, which act to change the function of transcription factors, have different inherited time scales. Genetics modification is a major cause of cellular memory inherited from parents to children. On the other hand, epigenetic modification plays a central role in cell memory acquired during a certain generation.

  However, the discovery of transgenerational epigenetics has shown that environmental epigenetics introduced into the germline resist developmental reprogramming and are inherited from parents to children (Stem Cell Res .4; 157-164, 2010). In addition, new genetic modifications such as mutation and translocation of DNA introduced into somatic cells are considered to be important factors for cancer formation.

  The "model of cell memory that evolves with time" is a concept newly introduced to formulate "cell memory that evolves with time" that changes over time due to genetics modification and epigenetic modification. Expression characteristics of transcription factors that change over time, genetics modifications, and epigenetic modifications are integrated as a “chromosomal state”.

  This is different from the conventional definition of two states in which cell memory is on / off for one transcription factor. In contrast, the chromosomal state of the present technology allows the transcription factor to induce and suppress the target gene product. Changes discretely. By introducing the concept of chromosomal state, the cell state is formulated as a concatenation of two time-series changes, the chromosomal state and the external environmental state of the cell. Furthermore, “time-evolving cell memory” is formulated as a function of time-series changes in “chromosomal state”.

  The "biological hierarchy connection model" is newly proposed to connect the "chromosome state" and "individual state" required for using "evolving cell memory" to predict individual health and disease. It is a model. A human individual is a hierarchically ordered continuum in which cells are composed of molecules, and more complex ones such as tissues and organs are constructed on the cells. Based on such a hierarchical structure, the human state can be classified into a general state, a local state, and a chromosomal state, and models of three different scales, macro, meso, and micro, can be constructed corresponding to each state. .

  The macro model is a model in which the whole body is regarded as one state. Therefore, the state can be defined by the expression characteristics of molecules that diffuse throughout the body. Typical examples of molecules that diffuse throughout the body are endocrine and immune system hormones, growth factors, and cytokines. In addition to this, adrenaline and noradrenaline, which control the autonomic nervous system, also regulate the general condition.

  The meso model is a model that captures the state of a homogeneous local space in a living body. Each organ or tissue forms a homogeneous state as a unit. Inflammatory symptoms generally occur locally and specifically. It is composed of a large number of cells that form the site of inflammation. On the other hand, tissue neurogenesis is controlled by proliferation and differentiation of a small number of tissue stem cells. Therefore, the meso model is composed of different numbers and types of cells such as one cell, inflamed tissue, one tissue, and organ. This local state is defined by systemic molecules in addition to the autocrine and paracrine molecule groups expressed in the local environment.

The micro model is a model that captures the state of chromosomes formed in each cell.
The reason why we do not make the micromodel a cellular state is that the cellular state can only represent the explicit cellular memory driven by transcription factors, and the potential changes consisting of genetic modification and epigenetic modification are discarded. . Chromosome status is defined by genetics modification, epigenetic modification properties, and transcription factor expression properties.

  Within the chromosomal state, epigenetic modifications and expression characteristics of transcription factors are controlled by the local and systemic environment of chromosomal cells. Cells that produce systemic molecules are also controlled by the local state and chromosomal state of the cells in addition to the systemic state, and cells that produce autocrine and paracrine factors, which are molecules of the local environment, are Controlled by environment and chromosomal status. That is, it became clear that the analysis formulates that the “chromosomal state”, the local state, and the general state together determine changes of each other, direct them together, and develop together with time. This formulation is called a "biological hierarchy connection model".

  Human beings are stimulated from various different environments depending on their social environment and their behavioral characteristics. These external stimuli are recognized by various systems of somatic cells and translated into changes in the internal environment of the human body. Therefore, stimuli from the same environment do not necessarily induce the same changes in the internal environment due to individual differences in somatic cell systems that receive the stimuli. This means that changes in the human body's condition cannot be accurately predicted only by quantitative measurement of environmental factors. Environmental stimuli are characterized by physical stress (temperature, oxygen, ultraviolet rays, etc.), chemical stress (endocrine disruptors, carcinogens, etc.), psychosocial stress, exercise stress, nutritional stress (overeating, starvation), infection. It can be classified into stress and injury stress.

  Environmental stimuli, after being sensed by the receptor system, are translated into changes in general, local, and "chromosomal status" with different time delays. Therefore, in order to develop "cell memory that evolves with time" into memory at the whole body level, it is necessary to capture time as a discrete distribution rather than as a continuous distribution. The time delay instability is controlled in vivo by a feedback system. Therefore, if a state change can be captured in a time unit such that a change from a certain feedback system to a new feedback system can be captured, the problem of time delay can be eliminated.

  The “living body state space model” is a model newly proposed in order to apply the “living body hierarchical connection model” to a time series model. The various models used in the time series model can generally be handled uniformly by a state space model (Introduction to time series analysis: Kitagawa Genshiro Iwanami Shoten, Chapters 9, 11 and 12). Many problems of time series analysis can be formulated as problems of state estimation of state space model.

  The state space model is composed of two submodels of a system model (x) and an observation model (y), and can be interpreted in two ways. Considering the observation model as a regression model that expresses the mechanism by which the observed time series data yn is observed, the state xn of the system model (x) becomes its regression constant. In this case, the system model is a model that expresses how the regression coefficient changes with time. On the other hand, considering the state xn as a signal to be estimated, the system model represents the generation mechanism of the signal, and the observation model represents that the signal is converted and noise is added when the signal is actually observed.

  In the “dynamically constructed model”, it is assumed that xn represents the actual state of the human body, yn is a multivariate time series vector regarding the expression levels of the extracellular factor groups that can be measured, and the state xn is the signal to be estimated. Think When applying the "biologically connected model" to clinical prediction, it is not realistic to measure all "chromosomal states" and local states. However, extracellular factors such as hormones, growth factors, and cytokines that circulate throughout the body can be measured.

  The "biological hierarchy connection model" revealed that the "chromosomal state", local state, and systemic state together determine changes in each other, direct them together, and substantially co-evolve. This means that the data of time-series changes of extracellular factors circulating in the whole body include changes of "chromosomal state" and local state. That is, the local state and “chromosomal state” components are directly or indirectly reflected in yn, which is multivariate time-series data consisting of systemic blood factors of human state xn. This formulation method is referred to as "biological state space model".

  The time-series data yn of extracellular factors circulating in the whole body are used as a periodic component by using a seasonal adjustment model, an environmental stimulus response component by using a multi-linear model, and a baseline component by using a polynomial smoothing spline model. , Each can be disassembled. Cellular memory is reflected in the amplitude and frequency of cyclic components, the maximum expression level of stimulus-responsive components, and long-term baseline changes related to the expression of molecules produced in the living body. Therefore, by combining the “biologically hierarchically connected model” and the “biological state space model”, it becomes possible to specifically represent the “time-developing cell memory” in a time series.

  "Dynamic construction model" is a control theory for a human body that evolves over time. One of the functions of biological evolution over time is that an initial stimulus from the environment alters its responsiveness to similar stimuli thereafter. The molecular entity of the change in reactivity is referred to herein as the "constructed change." With this formulation, it becomes possible to control the human body that evolves over time.

  Among the “structured changes”, the phenomenon that “the response to the next strong stress is improved when stress that is not lethal to the living body is improved” is generally called holmysis (Toxicology and Applied Pharmacology 222: 122-128, 2007). So far, the molecular mechanism of hormesis has not been clarified. In the “time-developing cell memory model” of this technology, it is clear that “constructed changes” such as holmysis have a molecular basis for dynamic cell memory by genetic modification and epigenetic modification. I chose This makes it possible to understand that the onset of chronic diseases is caused by the mismatch of the "time-developing cell memory" of the present technology with the subsequent stress.

  “Constituted changes” differ from the conventional control models, such as dynamic equilibrium and dynamic non-equilibrium models, which assume a fixed genetic algorithm based on genetic determinism. In other words, "constructed change" means that the function of a gene changes due to input from the environment or an accidental factor, and a part of the change is inherited in the future to change the genetically defined algorithm. It is assumed. Therefore, a new control theory is needed for "constructed changes".

In genetic determinism, the phenotypic temporal change (P (t)) is expressed as the product of a certain genotype (Gx) and an environmental factor (E (t)) that changes with time as follows. .
P (t) = Gx × E (t) (1)

On the other hand, in contrast to the "constructed change", the phenotypic time change (P (t)) is a chromosomal state (C (t)) that changes over time incorporating the "time-developing cell memory" of the present technology. ) And an environmental factor (E (t)) that changes with time, and is expressed as follows.
P (t) = C (t) × E (t) (2)
The biological model represented by Expression (2) is referred to as a “dynamically constructed model”.

  “Biological local stationary model” is a new data processing method for controlling health and disease using the “dynamically constructed model” of the present technology. Time-series data relating to the expression level of molecules produced in the living body are generally non-stationary. However, even if it is non-stationary, it can be assumed that each time interval is stationary by dividing the time interval into appropriate sub intervals. A time-series model in which a time interval can be divided into appropriate small intervals by calculation independently of the function of such a system and each small area can be assumed to be stationary is generally called a local stationary model (introduction to time series analysis). Genshiro Kitagawa Iwanami Shoten Chapter 8).

  In the "Biological Local Stationary Model", in order to allocate the state changes of cell memory as different steady states, the time series data of the expression level change of the molecule expressed in the cells in the living body or the cultured cells is first referred to as "the biological state space model". Is used to decompose the periodic component, the environmental stimulus response component, and the baseline trend component to identify the local constant region. Cellular memory influences variations in baseline components, amplitude and cycle of periodic components, and maximum and minimum expression of environmental stimulus response components. Among these, three categories that can be identified independently of the intensity of the input from the environment are the fluctuation of the baseline component, the amplitude of the periodic component, and the period.

  In the "biological local stationary model", a local stationary region is identified from these three categories. In general, the local stationary region can be classified into three types: an original stationary state, a new stationary state, and a change from the original stationary state to a new stationary state. The “biolocal stationary model” of the present technology is fundamentally different from the conventional local stationary model in that the stationary region is specified based on the characteristics of the organism, rather than simply mechanically decomposing the time series change into the stationary region. Different to

  The "method of tracking the biological state and analyzing the causal structure using the local constant region as a node" is to identify the internal factors that induce the state change in the human body that evolves over time, thereby maintaining health and developing disease. It is a method of anticipating and controlling. The disease begins at the subclinical stage of acquisition of potential changes, and becomes unsteady from acute symptoms, chronicity, some physiological function impairment, some physiological function loss, physical disability, bedridden, and death. Develops with time. That is, the time evolution of the disease can also be approximated as a continuous development of time segments that can be regarded as a steady state.

  In order to predict and control changes in the state of health and disease, it is necessary to first encode the local steady-state region of a disease that has a systemic and complex appearance, using molecules produced by cells as surrogate indicators. is there. This means that time-series data of molecules expressed from cells of organs, tissues, and subsystems associated with disease are divided from the viewpoint of local constant regions and associated with the disease phenotype. The types of molecules that are effective for this association are the optimal ones obtained by analyzing the data of the time-series changes in the expression levels of in-vivo molecules from many individuals by Bayesian method, using the causal structure model constructed from biological knowledge as a prior probability. Can be singled out. Therefore, among various molecules produced by cells, it is possible to relatively easily obtain time-series data from blood molecules using blood samples, and the application of the "dynamic construction model" of the present technology. Is effective for.

  As an instrument for measuring blood molecules, the glucose meter is considered as a prototype device (Freestyle Wright, Abbott). The glucose meter can measure the blood glucose level from a small amount of blood collected using a pain needle. This system can be applied to the measurement of blood molecules other than glucose. On the other hand, a method of transcutaneously measuring a molecule in blood has also been recently developed.

  The time segmentation of the constant region extracted from the time-series changes in the expression level of blood molecules not only is effective in tracking the changes in the states associated with various diseases, but also causes the onset of the disease. It can also be used to analyze causal relationships. The problem of health care is to identify changes in blood molecules that cause a transition from a healthy state to an asymptomatic early stage of the disease, and use it to prevent the onset of the disease. In addition, the problem of disease management is to identify changes in blood molecules that cause early disease into complications and to use the development of complications for prevention.

  In order to perform control using the “dynamically constructed model” of the present technology, the causal relationship that exists between changes in the expression level of biomolecules is first modeled using biological knowledge. Then, the constant region is extracted from the data of the time-series changes of the molecules in the living body of a plurality of patients, and is used to create the graph structure regarding the probability structure / causal structure between the identified nodes. After that, the time-series data of changes in the expression of biomolecules of a large number of healthy people and patients are used to improve the graph structure to an optimum one. Changes in the graph structure can be solved as a covariate selection problem in regression analysis. The constructed optimal graph structure is to predict and control changes in future biological conditions for the purpose of managing individual health and disease by continuing to measure changes in expression of biomolecules over time. Can be used for.

<First Embodiment>
[Application of dynamic construction model to prediction and control of health status and disease status]

  Developing effective preventive measures against diabetes, cancer, immune disorders, dementia, and cardiovascular diseases based on individual diversity and diverse past history is an important challenge for modern medicine. By using the “dynamically constructed model” of the present technology, individual prevention can be performed for these disease groups.

FIG. 1 is a block diagram showing the configuration of a biological information processing apparatus 1 that predicts and controls a health condition.
The biometric information processing device 1 includes a selection unit 11, a determination unit 12, a measurement unit 13, a creation unit 14, a division unit 15, an identification unit 16, and an estimation control unit 17.

  The selection unit 11 selects a measurement molecule. The determination unit 12 determines the molecular measurement interval. The measurement unit 13 performs measurement. The creation unit 14 creates a graph. The division unit 15 divides the component. The identification unit 16 identifies a constant region and a causal structure, and identifies a causal relationship between molecular markers and a relationship between external environments. The estimation control unit 17 estimates and controls the state change.

  FIG. 2 is a flowchart illustrating a process of predicting and controlling the health condition. Hereinafter, the prediction control process executed by the biometric information processing apparatus 1 in FIG. 1 will be described with reference to FIG. 2.

  In step S1, the selection unit 11 selects a measurement molecule for prediction and control related to the metabolic syndrome. That is, the type of blood molecule whose expression level is to be measured is selected.

  The “dynamically constructed model” of the present technology can be applied to predict and control the onset of a series of diseases that develop from metabolic syndrome to cardiovascular diseases. Diabetes ranks among a series of diseases that evolve from metabolic syndrome to cardiovascular disease (Nature 444; 839-888, 2006). Metabolic diseases cause obesity, visceral obesity, adiponectin insufficiency, insulin resistance, etc. due to lifestyle problems such as uneven diet and lack of exercise. The pathology then progresses to hypertension, postprandial hyperglycemia, and dyslipidemia. The stage up to this point is called the metabolic syndrome.

  Metabolic syndrome induces inflammation in visceral fat, activates the regeneration system of various organs, and raises oxidative stress. As a result, fatty liver and non-alcoholic steatohepatitis are caused. Then, diabetes occurs due to pancreatic dysfunction and insulin secretion deficiency.

  Diabetes induces nephropathy, retinopathy, neuropathy, arteriosclerosis obliterans, cerebrovascular disease, and ischemic heart disease, resulting in renal dialysis, blindness, lower limb amputation, stroke, dementia, and heart failure, leading to death. In particular, the inflammation induced by metabolic syndrome causes cancer and neurodegenerative diseases. When the “dynamically constructed model” of the present technology is applied to obesity and type 2 diabetes, the systemic blood factor shown below can be used as a molecular marker and time-series data of its expression level can be used ( Nature 444; 839-888, 2006).

  First of all, the molecular group related to the interaction between adipose tissue and the hypothalamic pituitary plays an important role in feeding, glucose metabolism, lipid storage, and energy balance. Can be used to represent and analyze. For example, chronic social stress has an epigenetic effect on the interaction between adipose tissue and the hypothalamic-pituitary-adrenal system, and changes the production of glucocorticoid and aldosterone from the adrenal gland. As a result, it alters appetite and sleep functions.

  In order to capture this change, in addition to glucocorticoid and aldosterone in the blood, time-series changes in the expression levels of ACTH produced by the pituitary gland, leptin produced by adipose tissue, adiponectein, visfatin, and omentin were measured. It is useful to To express and control obesity, visceral obesity, adiponectin dyssecretion, and insulin resistance, time-series changes in the expression levels of restin, RBP4 produced by adipose tissue, insulin produced by the pancreas, glucagon, amylin, GLP, etc. It is useful to measure From metabolic syndrome to inflammation in visceral fat, liver-produced CRP, PAI-1, NEFA, VLDL, LDL-ox, TNF-α produced from immune cells and adipose tissue, IL-6, MCP1 It is useful to measure the time-series change of the expression level of.

  In step S2, the selection unit 11 selects a measurement molecule for predicting and controlling an inflammatory disease.

  Immune diseases such as rheumatism, collagen disease represented by systemic lupus erythematosus (SLE), ulcerative colitis, and multiple sclerosis may develop due to past infection history and characteristics of intestinal bacteria. It has been known. From granulocytes, monocytes, macrophages, dendritic cells, natural killer cells, T cells that play a role in adaptive immunity, helper T cells (TH0, TH1, TH2), CD8 + positive T cells, regulatory T cells, and B cells Tracking changes in characteristic cytokines produced as molecular markers is useful for predicting and controlling immune system diseases. Examples of cytokines include interleukins, interferons, chemokines, growth factors and cytotoxic factors.

  In step S3, the determination unit 12 determines the molecular measurement interval.

  It is desirable that the time interval for measuring the expression level of the molecular marker is short, but depending on the measurement environment, time series data with long intervals such as days, weeks and months can also be used. Since many molecules fluctuate in expression level with a circadian rhythm of a 24-hour cycle, it is necessary to set a specific time in a day and acquire data when measuring at a time interval of 1 day or more. Is. Regardless of the time series data acquired, regardless of the time interval, it is desirable to analyze about 100 discrete data as one set. Therefore, in order to obtain the accuracy, it is required to measure at least 2-3 times a day, preferably about 50 to 100 times a day.

  In step S4, the measurement unit 13 executes measurement. Here, the expression level of a molecule in blood as a biological molecule is measured. Specifically, molecules that represent the general condition, the local condition, and the chromosomal condition are measured. This measurement is performed at the intervals determined in step S3.

  A pain needle can be used as a device for measuring a molecular marker in blood. A small amount of blood can be collected using this pain needle. Depending on the molecule, a method of transcutaneously measuring the molecule in blood is also possible. The type and amount of the molecule can be determined by fluorescently labeling an antibody selective for the molecule to be measured.

  In step S5, the creation unit 14 creates a time-series change graph. Specifically, graphs based on the equations (5) to (11) shown below are created. Further, in step S6, the division unit 15 divides the component. That is, three components are extracted from the time-series data using the biological state space model. The three components can be, for example, a periodic component, an environmental stimulus response component, and a baseline component.

  It is necessary to extract the necessary information for prediction and control from the time-series measurement data of molecular markers. As a method, it is desirable to handle multivariate time series data composed of a plurality of blood molecules in a unified manner as a "state space model". The "state space model" is composed of two sub-models, a system model and an observation model. Generally, these two sub-models are expressed using conditional probabilities as follows.

_xn- q ( _xn | _xn-1 ) (3) (system model)
y _n ~ r (y _n | x _n ) (4) (Observation model)

Where y _n is the observed time series of multivariate, x _n directly to k-dimensional vectors can not be measured, q and r, y given given by x _n-1 each x _n, and by x _n The condition distribution of _n is shown. In the “biological state space model” of the present technology, x _n represents the actual state of the human body, and y _n represents the measurable expression level of blood molecules. _Assuming that the distribution of the initial state vector x ₀ is arranged by P (x ₀ | y ₀ ), the prediction of the disease state is an evaluation of P (x _n | y _n ), that is, x given by the observation value y _n. To find the distribution of _n as well as the initial distribution.

  Of measurable blood molecules such as molecules that can be used to represent the process of progression from metabolic syndrome to diabetes to cardiovascular diseases, molecules related to immune system diseases, and endocrine molecules Each time series data y (t) is represented by the following equation. That is, it can be expressed by a model composed of three different components.

y (t) = s (t) + x (t) + b (t) + v (t) (5)
s (t): periodic component x (t): environmental stimulus response component b (t): baseline component v (t): observation error

The environmental stimulus response component x (t) can be formulated as a multi-linear model as follows.
x (t) = F (t ) x (t-1) + v x (t) (6)
i = 1, 2, 3, ..., N

In the above equation, F (t) represents the conversion function of the output with respect to the environmental stimulus, and v _x (t) represents the change of the environmental stimulus. Optimal values are selected for the parameters of the conversion function by applying the Yule Walker method, the least squares method, or the PARCO method to the autoregressive model.

ａ_tは回帰係数、ε（ｔ）はノイズ成分を示す。 The baseline component b (t) can be defined as a regression model of degree m by the polynomial smoothing spline model as follows.

a _t represents a regression coefficient, and ε (t) represents a noise component.

Further, the baseline component b (t) can be expressed as follows.
b (t) = H (t, t-1) b (t-1) + V (t, t-1) (8)
H (t, t-1) is an m × m matrix, and V (t, t-1) is a matrix relating to m-dimensional noise. This model is used to select an optimal function that makes the baseline smooth.

In the “dynamically constructed model” of the present technology, the main constituent factor of the periodic component s (t) is the circadian rhythm of the 24-hour period. For such a periodic component, when p observation values are obtained during one period, the following equation is approximately satisfied.
s (t) = s (tp) (9)

If this is expressed using the time-delayed operator G, the following equation is approximately satisfied.
(1-G ^p ) s (t) = 0 (10)
Therefore, the periodic component of order l can be formulated as the following seasonal adjustment model by the white noise ν (t).

  By this processing, the time series data of blood molecules can be divided into a periodic component, an environmental stimulus response component, and a baseline component.

  In step S7, the identification unit 16 identifies the constant region. That is, the stationary region is identified by using the “living body stationary model”.

  The local stationary region can be found by analyzing time series data with a model of periodic components, environmental stimulus response components, and baseline components. In the "dynamic construction model", it is useful to focus on the steady state change caused by "time-evolving cell memory" and the acquisition of a new steady state. Cellular memory influences variations in baseline components, amplitude and cycle of periodic components, and maximum and minimum expression of environmental stimulus response components. In this, it is the fluctuation of the baseline component and the fluctuation of the amplitude and period of the periodic component that can be identified independently of the intensity of the input from the environment. Focusing on this point, the local stationary region is identified from at least one of the fluctuation of the baseline component, the fluctuation of the amplitude of the periodic component, and the fluctuation of the period of the periodic component.

  In addition to the circadian rhythm of the 24-hour cycle, the time-series fluctuations of blood molecular groups include seasonal fluctuation patterns that repeatedly appear in units of months and years. Interpreting this as a baseline change can lead to false predictions and control strategies. Such seasonal components are extracted from the baseline components and divided as long-term periodic components.

  In step S8, the identification unit 16 identifies a causal relationship. That is, a causal relationship between the steady states of each molecule is created. A dynamic construction model can be used to identify the causal relationship.

  The constant regions of the molecular groups in blood related to “time-developing cell memory” extracted by this method form a direct or indirect causal relationship in time series. In order to apply this causal relationship to the control of disease progression, the causal structure is expressed by creating a graph structure related to the stochastic structure / causal structure using the constant region as a node.

FIG. 3 is a diagram showing a model regarding a causal relationship in which _four constant regions y _{1 to} y ₄ of five blood molecules A, B, C, D, and E are nodes. In FIG. 3, Y _i (A) to Y _i (E) are nodes of the constant regions y _i (i = 1, 2, 3, 4) of blood molecules A to E, respectively, and x _{0 to} x ₄ Represents the state of the living body. The states x ₀ and x ₁ represent a health state, the states x ₂ and x ₃ represent a symptom non-expression state, and the state x ₄ represents a symptom development state. The numbers represent the passage of time going from a smaller value to a larger value.

In this model, node Y ₄ (A) and node Y ₄ (D) are both markers and substitutes for disease states. A node Y ₃ (A) and a node Y ₃ (C) work to guide the node Y ₄ (A), and a node Y ₃ (E) works to guide the node Y ₄ (D). Similarly, the nodes Y ₂ (C), Y ₂ (A), and Y ₂ (D) work to guide the nodes Y ₃ (A), Y ₃ (C), and Y ₃ (E), respectively. Further, nodes Y ₁ (A), Y ₂ (C), and Y ₂ (D) are guided by nodes Y ₁ (A), Y ₁ (B), and Y ₁ (D), respectively.

  The estimation of the graph structure regarding the causal structure between the constant regions of the biomolecule can be formulated as follows. First, we consider the steady-state value of biomolecule Molecule 1 as a random variable. That is, under some circumstances, the quantity of Molecule 1 is considered to be the realization value of the random variable X 1. The observation data processed by the biological local stationary model of the present embodiment can be regarded as data obtained by measuring the realized values of the random variables of p types of blood molecules under a certain situation. Since the steady state is observed at various times and situations, the observation data processed by the local steady model can be expressed as a data matrix. That is, the observation data composed of n different steady states has a matrix size of p × n. The purpose of this analysis is to estimate the dependence between the matrix data.

  Various mathematical models such as a Boolean network, a Bayesian network, a graphical Gaussian model, and an ordinary differential equation have been proposed as a method for estimating a graph representing a dependency relationship between random variables from a data matrix, which should be used in this embodiment. You can Among them, by assuming that there is a dependency represented by an acyclic directed graph among random variables, and by assuming a Markov chain rule for the structure of this acyclic directed graph and the dependency between nodes, the joint probabilities are The estimation by the Bayesian network represented by the product of conditional probabilities under the variable given its immediate parent random variable set is useful in this embodiment.

  For more detailed adaptation of Bayesian networks, the method used for data analysis of DNA chips can be applied (Yoshinori Tamada, Kiyoya Imoto, Satoru Miyano Vol. 54, No. 2, 333-356). Specifically, it is carried out by making the steady-state expression level of this embodiment correspond to the mRNA expression level of the DNA chip. The current physiological model can be used as prior knowledge for the steady-state graph structure estimation of the present embodiment.

  The graph structure is first constructed at the individual level. Since the current physiological models do not adequately support individual diversity and diversification, there is a possibility that graph structure estimation will be erroneous. To optimize the graph structure, a number of individual graph structures are compared, sorted into small groups with similar structures, and the graph structure is reselected. This builds an optimal graph structure.

  The causal route to common symptoms in many patients can vary from individual to individual. Therefore, the optimum graph structure is not converged into one, but the optimum graph structure is determined for each group having the same causal structure.

  In step S9, the identification unit 16 identifies the relationship between the causal relationship of the molecular markers and the external environment. That is, a causal relationship between each molecule and an environmental relationship are created. As the external environment, measured values of time-series changes in lifestyle can be used.

  When the input from the environment can be measured quantitatively, it is possible to analyze how the stimuli of different intensities received at different times affect the reactivity of blood molecules, and thus the “constructed changes” are analyzed. Can be applied to control. For example, the negative stress level can be measured by DRM (Daily Reconstructive Method). In this, one day is divided into about 24 on average for each event, and each positive and negative value is expressed step by step with a numerical value from 0 to 6.

  The effect of a meal can be measured by electronically recording the content of each meal, so that intake information such as calories, carbohydrates (sugars), fats, proteins, vitamins, and phytochemicals can be measured. It is also possible to measure the intake of supplements and the like. If a correlation is observed between these environmental stimuli and changes in the expression level of molecules in the blood, it can be used to control the changes in the expression levels of blood molecules for the purpose of the environmental stimulation, that is, changes in the state of the body. it can.

  In step S10, the estimation control unit 17 estimates and controls the state change. That is, future changes in the state of the human body are estimated, and intervention methods including preventive methods are proposed.

  The state change is estimated by classifying the population according to the similarity of the causal structure between the constant regions obtained in step S8. This makes it possible to estimate future changes in the disease preliminary group classified into the same group from the causal structure acquired from the patient with advanced disease. In addition, the causal structure in which a potential state before the onset of the disease is restored to a healthy state without the onset of the disease can be used as a control method for preventing the onset of the disease. For control, the causal relationship of the molecular markers (that is, the internal environment) identified in step S9 and the external environment relationship can be used.

<Second Embodiment>
[Application of dynamic construction model to standardization of cultured cells]

  FIG. 4 is a block diagram showing the configuration of the biological information processing apparatus 51 that predictively controls the cell state of cultured cells. The biological information processing apparatus 51 includes a selection unit 61, a measurement unit 62, a determination unit 63, a division unit 64, an identification unit 65, and an estimation control unit 66.

  The selection unit 61 selects a measurement molecule. The measurement unit 62 executes measurement. The determination unit 63 determines the measurement interval. The dividing unit 64 divides the components. The identification unit 65 identifies a constant region, a causal structure between molecular markers, and a causal structure between molecular markers and environmental conditions. The estimation control unit 66 estimates and controls the cell state of the cultured cells.

  FIG. 5 is a flowchart illustrating a process of predicting and controlling the cell state of cultured cells. Hereinafter, with reference to FIG. 5, the prediction control process executed by the biological information processing apparatus 51 in FIG. 4 will be described.

  In step S31, the selection unit 61 selects a measurement molecule.

  Changing the properties of cultured cells with culturing is a major obstacle in constructing an evaluation system using cells and developing therapeutic agents containing cells as active ingredients. The prediction and control of changes in the state of cultured cells can be formulated as the problem of mesoscale state estimation. The mesoscale state is expressed by the components added to the medium as well as the local extracellular environment formed by molecules produced from the cultured cells. Therefore, in order to predict and control the state of the cultured cells, the time series changes of the molecules produced in the cultured cells and the molecules in the medium are measured.

  Molecules produced from cultured cells include not only secretory molecules but also molecules expressed on the cell surface, but not intracellular signaling molecules or transcription factors. It is useful to measure the amount of insulin or IGF-I as a component in the medium. In addition to insulin and IGF-I, growth factors such as FGF-2, EGF and PDGF, and integrin family molecules are useful as molecules produced by cells.

  In step S32, the measurement unit 62 executes measurement.

  Molecules produced from cultured cells as well as molecules in the medium are measured by obtaining a smaller sample than in the medium. In addition, among molecules produced from cells, those that are captured by extracellularly expressed molecules and do not float in the medium are quantified by fluorescently labeling the molecules and then performing image analysis of cultured cells. For molecules that are difficult to measure, the intracellular reporter system is used to perform alternative measurement by gene expression of secretory proteins or membrane proteins.

  In step S33, the determination unit 63 determines the measurement interval.

  It is desirable that the measurement interval be about 10 minutes to 1 hour. When a reporter system is used, data will be acquired continuously.

  In step S34, the division unit 64 divides the component. Further, in step S35, the identification unit 65 identifies the constant region.

  In order to derive information from the acquired time series data of molecular markers, multivariate time series consisting of molecules produced from cultured cells and molecules in the medium are treated as a state space model in a unified manner. The state space model is composed of two submodels, a system model and an observation model. Generally, these two sub-models are expressed using conditional probabilities as follows.

_xn- q ( _xn | _xn-1 ) (12) (system model)
y _n ~ r (y _n | x _n ) (13) (Observation model)

Where y _n is the observed time series of multivariate, x _n directly to k-dimensional vectors can not be measured, q and r of y _n which is given by x _n and x _n is given by x _n-1, respectively The condition distribution is shown. In the “biological state space model”, x _n represents the actual state of the cultured cells, and y _n represents the expression level of the measurable molecule in the medium and the molecule produced by the cultured cells. _Assuming that the distribution of the initial state vector x ₀ is arranged by p (x ₀ | y ₀ ), the prediction of the state of the cultured cells is given by the evaluation of p (x _n | y _n ), that is, the observed value y _n. It can be formulated by obtaining the x _n distribution as well as the initial distribution.

The time series data y _n of the molecular group produced from the cultured cells are used for the periodic component by using the seasonally regulated model, the environmental stimulus response component by using the multiple linear model, and the baseline by using the polynomial smoothing spline model. It is possible to decompose into components and extract. The cell memory of cultured cells is reflected in the amplitude and frequency of cyclic components, the maximum expression level of environmental stimulus response components, and changes in baseline. From this, the local stationary region related to the change of the amplitude and frequency of the periodic component and the baseline is noted as "time-developing cell memory".

  In step S36, the identification unit 65 identifies a causal relationship between molecular markers. In step S37, the identifying unit 65 identifies the causal relationship between the molecular marker and the environmental condition.

  Changes in amplitude and frequency of periodic components, changes in baseline, temporal changes in artificial components added to the medium, and environmental stimulus response components form a direct or indirect causal relationship according to a time series. In order to apply this causal relationship to the control of changes in the state of cultured cells, the graph of the stochastic / causal structure is created using the local steady states of the molecules in various media and the molecules produced by the cultured cells as nodes. The causal structure is expressed by making it.

  After creating the graph structure based on the conventional biological knowledge, it is improved to the optimum graph structure using time series data. Changes in graph structure are analyzed as a covariate selection problem in regression analysis. Depending on their origin, cultured cells may have different causal pathways leading to the same state change. Therefore, the optimum graph structure is not converged into one, but the optimum structure is determined for each population of cultured cells having the same causal structure. In classifying this group, the state change characteristics of the "biological state space model" can be used. In addition, a causal structure for intervention by changing environmental conditions such as medium components, cell culture density, oxygen partial pressure, and temperature will be created separately.

  In step S38, the estimation control unit 66 estimates and controls the cell state of the cultured cells.

  The estimation and control of changes in cell state due to culturing can be carried out from the past causal structure obtained by culturing cells derived from the same tissue under similar conditions. In addition, control toward the direction of obtaining the state of cells can be carried out from changes in environmental conditions such as medium components, cell culture density, oxygen partial pressure, temperature, and the causal structure of changes in cell states.

<Third Embodiment>

Next, a model of cells that produce molecule A secreted in blood will be examined. This cell has the following characteristics, as shown in FIG.
(1) It is controlled by the circadian rhythm in the steady state.
(2) Induced expression in response to a stimulus from the environment.
(3) Receptor of molecule A and repressor R receive positive and negative feedback.
(4) The expression of repressor R is suppressed by epigenetic modification.

The time series change of molecule A secreted from this cell into blood changes like y in FIG. 7A.
This time series measurement data y (t) is assumed as follows.

y (t) = s (t) + x (t) + b (t) + v (t) (14)
s (t): periodic component x (t): environmental stimulus response component b (t): baseline component v (t): observation error

The environmental stimulus response component x (t) was formulated as a multiple linear model as follows.
x (t) = F (t) x (t-1) + v (t) (15)

  In the above formula, F (t) represents the conversion function of the output of the cell with respect to the environmental stimulus, and v (t) represents the environmental stimulus. The parameters of the conversion function were searched for the optimum one using an autoregressive model.

The baseline component b (t) was defined and used as a regression model of degree 2 as follows.
b (t) = a _t-1 b (t-1) + a _t-2 b (t-2) + v _B (t) (16)
Here, a _j is a regression coefficient and v _B (t) is a noise component. Then, b (t) can be expressed as follows.
b (t) = H (t, t-1) b (t-1) + V (t, t-1) (17)

  Here, H (t, t−1) is a 2 × 2 matrix, and V (t, t−1) is two-dimensional noise. The function was optimized so that the baseline was smoothed.

Since the periodic component s (t) is based on the circadian rhythm of a 24-hour period, when p observation values are obtained during one period, the periodic component is s (t) = s (tp ) (18)
Will be approximately satisfied. If this is expressed using the time-delayed operator G, the following equation is approximately satisfied.
(1-G ^p ) s (t) = 0 (19)

Therefore, the periodic component of order l can be formulated as the following seasonal adjustment model by the white noise ν (t).

  Using this model, the time series data of y in FIG. 7A was decomposed into a periodic component s, an environmental stimulus response component x, and a baseline component b. In addition, i in FIG. 7A represents an induction component.

Next, it was decomposed into three constant regions Y ₁ (A), Y ₂ (A), and Y ₃ (A) based on the variation characteristic of the baseline. The cell has an input from the environment between times k and l, and the environmental stimulus response component of Y ₂ (A) is a direct output to this input. Further, in the present self-regulation model, the time change segment Y ₂ (A) of the molecule A is the cause of the change of the expression state of the factor A from Y ₁ (A) to Y ₃ (A). This relationship is shown as a change from Y ₂ (A) to Y ₃ (A) as shown in FIG. 7B by the graph structure in which the constant region relating to the change in the expression amount of the A molecule is noded.

  As described above, in each of the embodiments, changes over time in the human body can be formulated, and changes in the body condition that can be used for health management, disease prevention, disease cure, etc. can be predicted and controlled.

  Further, it can be applied to control the onset of chronic diseases caused by long-term lifestyle habits and the progression of chronic diseases, which cannot be handled by conventional methods. Furthermore, since it can handle commonality and heterogeneity among individuals at the same time, it is an effective tool for individualized medicine. In addition, it is possible to control the diversification of cells induced by cell culture due to time development, and it can be applied to standardization of cell therapy and cell evaluation system.

<Fourth Embodiment>

  Many children born underweight due to fetal failure to develop due to fetal malnutrition show growth that can be backed up by the age of two, and thus are of the same size as children born to normal weight. When a low-calorie diet is given as a diet after birth, the growth that catches up with the delay is suppressed, and even after the age of two, the growth cannot be regained by the high-calorie diet. The risk of observing insulin resistance in the future is increased if there is rapid growth to catch up with this delay (Rotteveel, J. et al. European Journal of Endocrinology 158; 899-904, 2008). Even if the physical constitution is regained as a cause, the differentiation efficiency of β cells of the pancreas changed by epigenetics to adapt to the nutritional environment in the womb, insulin signal responsiveness, appetite level, stress responsiveness, etc. are regained. It is speculated that there is no such thing.

  However, not all underweight infants with fetal failure to show growth to catch up develop insulin resistance. Insulin secretion ability, insulin signal responsiveness, appetite level, and stress responsiveness of each person do not show rapid changes such as growth to catch up with delay, but since they change depending on lifestyle, blood insulin level, blood insulin level By monitoring the amount of leptin, the amount of glucocorticoid in blood, the blood glucose level, the amount of calorie intake, etc. using the method of the present embodiment, it becomes possible to predict and control obesity tendency, insulin resistance onset tendency, etc. .

  Specifically, by measuring and comparing the relationship between the calorie intake by meals, the amount of insulin secretion after meals, and the time change of blood glucose level every day, the amount of insulin secretion produced per calorie intake and the change in the strength of insulin responsiveness can be determined. , It is possible to measure the change from one steady state to another steady state. Steady-state changes can be classified into changes that improve insulin resistance and changes that worsen insulin resistance. An increase in the amount of insulin secretion as well as the intensity of insulin responsiveness leads to a change in the improving direction, and vice versa.

  The amount of glucocorticoid or leptin secreted can also be formulated for the first time as a change from one steady state to another steady state by using the method of the present embodiment. Low leptin production for a given caloric intake weakens food intake and increases obesity. High glucocorticoid level increases obesity tendency and cardiovascular disease onset tendency through multiple sites of action, leading to a decrease in glucocorticoid secretion and a change in steady state in the direction of increasing leptin secretion. Is a change in the direction of improving insulin resistance and the tendency of developing related diseases.

  One of the features of this embodiment is that it tracks a wide range of temporal changes of various molecules in a living body. Therefore, the present embodiment can also be used to identify an unknown biological molecule involved in changing insulin secretion, insulin responsiveness, glucocorticoid secretion, and leptin secretion from one steady state to another steady state. . Further, by clarifying the relationship between the record of lifestyle and the novel related biomolecule using the method of the present embodiment, it is possible to propose a new improving method for improving insulin resistance and obesity. Become.

<Fifth Embodiment>

  The blood level of DHEA-S, in which sulfuric acid was added to dehydroepiandrosterone, began to increase from around age 6 to 7 in both men and women, reached a peak around age 12 to 13, and reached a high value around age 13 to 25. Maintains and then decreases linearly with aging. Long-term follow-up studies have shown that blood DHEA-S levels in men are inversely correlated with mortality and the development of cardiovascular disease in multiple studies around the world (Toshihiko Yanase, Anti-Aging Medicine vol5 No. 42-46, 2009). However, it is not currently clear what reduces production of DHEA-S. According to the present embodiment, the biomolecule involved in changing the secretion amount of DHEA-S from one steady state to another steady state and the lifestyle related to the fluctuation of the biomolecule are clarified, and the decrease of DHEA-S is suppressed. It becomes possible to make a new proposal for a suppression method.

<Sixth Embodiment>

  The most effective way to delay aging and prolong life is calorie restriction. Calorie restriction is a calorie restriction that is carried out while taking optimal nutrition, rather than simply reducing the amount of food. However, there is insufficient scientific evidence as to what is optimal nutrition or to what level calorie restriction should be done. The present embodiment can be used to clarify the optimum nutritional intake level and caloric intake level at the individual level.

  For example, when calorie restriction is given to patients in the diabetic army, the time series changes of adiponectein production, which is an index of qualitative changes of adipocytes, active oxygen production, and c-reactive protein (CRP), which is an index of inflammation Can be measured. As a result, the state of fat cells, active oxygen stress state, and inflammatory state can be formulated as changing from one steady state to another steady state.

  The present embodiment can also be used for identifying a new biomolecule related to the change from the steady state to another steady state. Time-series recordings of newly discovered biomolecular markers and diets can be used to individually optimize nutritional intake as well as caloric intake.

<Seventh Embodiment>

  The difference in sensitivity to stress induces the onset of depression and suicide associated with chronic stress. Susceptibility to stress depends on the strength of negative feedback by glucocorticoid-responsive neurons that project from the hippocampus to the amygdala. In addition, the intensity of negative feedback of the neuron depends on the epigenetic modification of the glucocorticoid receptor gene. Comparing epigenetic modifications of the neurons in the suicide and control groups shows that those who have undergone child abuse or child abandonment have their glucocorticoid receptor genes suppressed by epigenetic modifications. (Hyman, S. Nature Neuroscience Vol. 12, No. 3, 241 243, 2009).

  Understanding the level of stress sensitivity can prevent the onset of depression by quantitatively ascertaining the person who is likely to develop depression and providing an appropriate environment. When the time-series change of blood glucocorticoid is measured at a short interval of about 10 minutes using this embodiment, the time-series change can be divided into a diurnal variation component and a stimulus-dependent component. From the diurnal fluctuation component, the average amount of blood cholesterol secretion on the measurement day can be calculated. In addition, the relative strength of negative feedback can be calculated from the stimulus-dependent component by measuring the maximum secretion amount and duration. Therefore, by applying this embodiment, it is possible to measure stress susceptibility, and it is also possible to trace the process in which stress susceptibility changes with lifestyle.

  The most well-known hypothesis regarding the mechanism of depression is due to a decrease in the neurotrophic factor BDNF (Brain Derived Neurotrophic Factor) (Shi, Y. et al. Psychiatry and Clinical Neuroscience 64, 249-254, 2010). BDNF is produced from pro-BDNF (BDNF precursor) by t-PA (tissue type plasminogen activator). That is, utilizing the present embodiment, by observing the time-series changes of BDNF, pro-BDNF, t-PA in blood, and identifying the change from one steady state to another steady state, the onset of depression. It is possible to estimate the therapeutic effect. This embodiment can also be used to find new biomolecular markers related to the onset and prevention of depression.

<Eighth Embodiment>

  Chronic overwork and mental stress, such as caring for a spouse, increase the frequency of infections, weaken the ability to heal wounds, and increase the frequency of hypertension and heart disease. As a biomarker related to such chronic overwork and mental stress, there is the inflammatory cytokine IL-6 (interleukin 6). It is also known that IL-6 is associated with cardiovascular diseases, osteoporosis, type II diabetes, cancer, periodontal disease, weak constitution, organ failure, etc. that develop in an age-dependent manner. IL-6 is induced by experiences with depression, negative emotions, and stress. In addition, the increase in blood IL-6 with age is significantly increased in those who are engaged in care compared with those who are not involved in care (Kiecolt-Glaser JK et al. Proc. Natl. Acad. Sci. USA 100: 9090-9095). Therefore, by tracking changes in the expression level of IL-6, it is possible to predict and prevent the onset of diseases associated with aging and stress.

<Ninth Embodiment>

  The human body is an open system, and the future of physical conditions is not just a repetition of the past. The cell memory of the present technology makes it possible to capture and quantify a process in which a living body evolves with time according to a history only by measuring biomolecules. In order to apply this quantified change over time to personalized healthcare, it is necessary to provide appropriate future lifestyle options based on individual diversity and diversification.

  Further, in order to estimate the past diversity and the diversification entity before starting the biological state measurement of an individual based on the present technology, the DNA sequence information inherited from an individual by their parents is useful. The current diversity of each individual is formed by genetic diversity and past lifestyles. Therefore, the accuracy of the future prediction of health can be improved by integrating the cell memory information of biomolecules, the DNA polymorphism information, and the environmental information based on lifestyle habits of the present technology.

Advances in nucleotide sequencing technology have made it possible to analyze individual whole genome sequences.
From such sequence analysis, it is possible to detect individual single nucleotide polymorphisms and structural polymorphisms. The relationship between genetic diversity and health or disease can be classified into four types: chromosomal abnormalities, monogenic diseases, multifactorial genetic diseases, and idiopathic diseases (Thompson & Thompson "Genetic Medicine 7th Edition" Medical・ Science International 2009).

  Human somatic cells consist of 23 to 46 chromosomes, 22 of which are autosomal sex-free. The remaining pair is called the sex chromosome. Females have two X chromosomes and males have one X chromosome and one Y chromosome. The frequency of chromosomal abnormalities is high, affecting about 6 in every 1000 newborns. The most common chromosomal abnormality is an abnormality in which the number of chromosomes increases. An increase in the number of chromosomes by one is called trisomy. For example, trisomy on chromosome 21 develops a disease called Down syndrome. About 1 in 800 is born with Down's syndrome.

  The frequency of births with Down's syndrome increases exponentially with the age of the mother, and approximately 45% of children over the age of 45 are born with Down's syndrome. The main problem with Down's syndrome is delayed mental development. However, other than that, the risk of leukemia increases, and many patients with Down's syndrome develop Alzheimer-type dementia after age 40. 90% of the causes of trisomy 21 are in the process of meiosis that forms the maternal egg. Therefore, the abnormality called trisomy of chromosome 21 exists from the moment of fertilization. However, the symptoms appear at various times, and there is a "time delay" between the establishment of the genetic cause and the onset of the disease.

  A monogenic disease is a disease that develops due to a mutation of a gene on the genome and is called Mendelian inherited disease. To date, 3917 types of Mendelian genetic diseases have been reported. There are dominant and recessive forms of inheritance. Dominance is a form of inheritance that develops only by inheriting a gene mutation from either the mother or father. On the other hand, recessive means a hereditary form that develops only when inheriting a gene mutation from both parents. It is said that about 2% of the population will find monogenic disease at some point during their lives.

  Monogenic diseases are diseases that mainly occur in childhood, but about 10% or less of monogenic diseases are symptomatic after puberty. Although it is less than 1%, it may occur in old age after the end of the reproductive period. Whatever the onset of life is, there is a "time lag" between the acquisition of the mutation and the onset of symptoms.

  Chronic diseases such as congenital abnormalities observed at birth, myocardial infarction, cancer, diabetes, rheumatism, mental illness, dementia, and other middle-aged and elderly diseases are called multifactorial diseases. The frequency of appearance of these types of diseases at birth is about 50 per 1000. However, since many people are suffering from chronic diseases, the number of people in a group is about 600 per 1,000. It is thought that the onset of disease is caused by a combination of multiple genetic factors, and that it involves daily and accidental exposure to certain environmental factors. The genetic effects of such multifactorial diseases have been analyzed by a method called genome-wide association analysis (GWAS).

  Idiopathic diseases are located between Mendelian mutations and multifactorial diseases. Parkinson's disease and Alzheimer's disease have sudden types other than familial ones, and it is necessary to use individual genome sequence analysis to identify such sudden mutations, unlike multifactorial diseases.

  Whether you have a chromosomal abnormality, a Mendelian monogenic disorder, a multifactorial disorder, or an episodic disorder, there is some time delay between the acquisition and onset of the mutation. Further, the explanation that a plurality of environmental factors cause the disease as an immediate response reaction is not appropriate. There is also a "time delay" between the time genetic factors and environmental factors are gathered until the onset of disease. In other words, genetic factors and environmental factors cause latent changes in the living body as cell memories, and become manifest as symptoms when they form a certain combination.

  Since cellular memory is acquired by changing environmental conditions on top of genetic polymorphism, quantitative recording of environmental conditions is effective for predicting future health. Among the environmental conditions, stress, dietary habits, exercise habits, infection history, treatment history, and other external environments can be quantitatively measured.

  Stress means that an individual recognizes that he / she is not adapted to the environment, and whether or not he or she feels stress under the same environmental conditions varies from person to person. Therefore, stress cannot be understood only from environmental conditions. It is effective to record stress by, for example, the Daily Reconstruction Method (DRM method) (Kahneman D, Krueger AB, Schkade DA, Schwarz N, Stone AA. `` A survey method for characterizing daily life experience: the day reconstruction method. Science. 306: 1776-1780, 2004).

  This method divides a day's event into about 24, and records the positive and negative values of each. Thereby, the situation of stress coping at each moment can be grasped. By continuing such records, stress can be classified into short-term stress of several hours to one day, medium-term stress of several days to one month, and long-term stress of several months or more. .

  Regarding dietary habits, by recording the dietary content by dividing it into ingredients, calories, carbohydrate ratio, protein ratio, fat type and amount, vitamins, phytochemicals, trace metals (iron, zinc, copper, cobalt, iodine, selenium) , Manganese, molybdenum, chromium, boron, vanadium), drinking (alcohol) etc. can be recorded.

  Regarding the amount of exercise, it is possible to automatically measure the quality and amount of walking and running by using a portable device having an acceleration sensor.

  The infection history and the treatment / medicine history can be grasped by the electronic medical chart data of the doctor. In addition, it is possible for individuals to record Kampo medicines and general medicines that they take.

  As described above, individual genetic diversity should be classified using single nucleotide polymorphisms and structural polymorphisms in DNA that cause chromosomal abnormalities, monogenic diseases, multifactorial genetic diseases, and idiopathic diseases. Is possible. Changes in environmental conditions with time can be quantitatively described by stress status, dietary intake component, exercise amount, infection history, and treatment / medication history.

In order to integrate the cell memory information of biomolecules, the DNA polymorphism information, and the environmental information based on lifestyle habits of this technology, we use the relation between the following three subgroups and conditional probability. To clarify. The three small groups are a small group divided based on genetic diversity, a small group divided based on changes in environmental conditions, and a small group divided based on cell memory information of biomolecules. Subpopulations divided based on genetic diversity include genetic variations such as single nucleotide polymorphisms and structural polymorphisms that cause chromosomal abnormalities, monogenic diseases, multifactorial genetic diseases, and idiopathic diseases. It is a group based on. The small group divided based on changes in environmental conditions is a group based on environmental conditions such as stress status, dietary habits, exercise habits, infection history, and treatment / medicine history.
The small group divided based on the cell memory information of the biomolecule is a group based on the change of the biological state represented by FIG. 3, which is created by the process shown in the flowchart of FIG.

Such conditional relationships of the group can be roughly classified into the following three, as shown in FIG.
i) Relationship between genetic diversity and environmental conditions
ii) Relationship between environmental conditions and biological condition
iii) Relationship between genetic diversity and biological condition

  The application of this technology to personalized health care means to maintain health based on past cell memory data of health care subjects, which is shown as a causal relationship between the constant regions of expression levels of biological molecules. Is to provide future options. As a method of finding such future options, the lifestyle of other people (that is, other than the target person) who have a history similar to the target person, such as the history of genetic information, the history of environmental conditions, the history of biological conditions, etc. It is possible to refer to the record of. This method can be carried out by clarifying the similarity in the correlation between cell memory information of individual biomolecules, genetic information such as DNA polymorphism information, and environmental information based on lifestyle shown in FIG.

  Such data similarity can be determined by applying an existing clustering method. The clustering method is large and can be classified into divided clustering and hierarchical clustering. Specifically, AK Jain, MN Murthy and PJ Flynn, Data Clustering: A Review, ACM Computing Reviews, (1999). And Ying Zhao and George Karypis, “Hierarchical Clustering Algorithms for Document Datasets”, Data Min. Knowl. Discov. . 10 (2): The method described in pp.141-168 (2005) can be applied.

  When performing the processing of this embodiment, the biometric information processing apparatus 101 is configured as shown in FIG. The biometric information processing apparatus 101 includes a selection unit 111, a determination unit 112, a measurement unit 113, a creation unit 114, a division unit 115, an identification unit 116, a detection unit 117, a search unit 118, a database 119, and an estimation control unit 120. is doing. The functions of the selection unit 111, the determination unit 112, the measurement unit 113, the creation unit 114, the division unit 115, the identification unit 116, and the estimation control unit 120 are the same as those of the selection unit 11, the determination unit 12, the measurement unit 13, and the measurement unit 13 illustrated in FIG. It is basically the same as the creation unit 14, the division unit 15, the identification unit 16, and the estimation control unit 17. Therefore, the description will be omitted because it is repeated.

  The detection unit 117 detects genetic information. The search unit 118 searches the database 119 for similar histories.

  The database 119 classifies and stores many pieces of personal information into any of the three small groups described above with reference to FIG. 8. That is, the information on the inspection results of many people obtained by performing the above-described processing of FIG. 2 is stored. Also, at this time, a measurement / inspection is added to obtain information necessary for determining which of the above-mentioned three small groups to classify. Specifically, in addition to the measurement and inspection of environmental conditions such as stress status, dietary habits, exercise habits, infection history, treatment and medication history, and measurement and inspection of cell memory information of biomolecules, single nucleotide polymorphisms in DNA In addition, measurements and tests to identify genetic diversity such as structural polymorphism will be added.

  The process of this embodiment is shown in the flowchart of FIG. The processing of this embodiment will be described below with reference to FIG.

  A database generation process is executed in step S101. That is, the process shown in the flowchart of FIG. 2 is performed on a large number of individuals together with the measurement / test for identifying the genetic diversity such as single nucleotide polymorphism and structural polymorphism of DNA of each individual. Are stored in the database 119. However, information is stored in a format that conceals the privacy of each individual.

As described above, the information stored in this database 119 is classified into a group based on genetic diversity, a group based on changes in environmental conditions, and a group based on cell memory information of biomolecules. Then the following three relationships in each population are analyzed using conditional probabilities.
i) Relationship between genetic diversity and environmental conditions
ii) Relationship between environmental conditions and biological condition
iii) Relationship between genetic diversity and biological condition

  Note that, for convenience of explanation, the database generation process is described as one step of the health condition prediction process of the flowchart of FIG. 10, but if the database has already been generated by some method and exists, it can be used. In this case, this database generation process can be omitted.

  The processing of steps S102 to S109 of FIG. 10 is basically the same as the processing of steps S1 to S8 of FIG. Again, these are briefly described. That is, in step S102, the selection unit 111 selects a molecule to be measured by a health care target person, and in step S103, the determination unit 112 determines the molecular measurement interval. In step S104, the measurement unit 113 measures the measurement molecule selected in the process of step S102 at the intervals determined in the process of step S103.

  In step S105, the creation unit 114 creates a graph of time-series changes. That is, the graphs based on the above equations (5) to (11) are created. Further, in step S106, the dividing unit 115 divides the component. That is, the periodic component, the environmental stimulus response component, and the baseline component are divided from the time series data using the biological state space model.

  In step S107, the identification unit 116 identifies the constant region. That is, the stationary region is identified by using the “living body stationary model”. In step S108, the identification unit 16 identifies a causal relationship. That is, a causal relationship between the steady states of each molecule is created. A dynamic construction model can be used to identify the causal relationship.

  In step S109, the identification unit 116 identifies the causal relationship between the molecular markers and the external environment. That is, a causal relationship between each molecule and an environmental relationship are created. As the external environment, measured values of time-series changes in lifestyle can be used. These measurement values are input by the subject himself or from a measuring device or the like.

  Through the above processing, cell memory information and environmental information of the subject's living body are acquired.

  In step S110, the detection unit 117 detects a DNA polymorphism. That is, the detection unit 117 detects a single nucleotide polymorphism and a structural polymorphism as genetic information of a gene possessed by a health care subject.

  When the history of the target person is obtained as described above, next, in step S111, the search unit 118 searches the information stored in the database 119 for the history of another person similar to the history of the target person. . That is, as described above, the similarity in the correlation between cell memory information of individual biomolecules, genetic information such as DNA polymorphism information, and environmental information based on lifestyle is determined, and the most similar history is searched.

Retrieval of similar histories will be further described with reference to FIGS. 11 and 12. For example, as shown in FIG. 11, it is assumed that five blood molecules A to E of a health care subject are selected and measured according to their purposes. Similar to the case in FIG. 3, FIG. 11 is a diagram showing a model regarding a causal relationship in which four constant regions y _{11 to} y ₁₄ of five blood molecules A, B, C, D, and E are nodes. Y _i (A) to Y _i (E) are nodes of the constant region y _i (i = 11, 12, 13, 14) of blood molecules A to E, respectively, and x _{10 to} x ₁₄ are It represents the state. State x _10, x ₁₁ represents the health, condition x _12, x ₁₃ represents the symptoms unexpressed state, the state x ₁₄ represents the manifestations state. The number represents the passage of time from the smaller value to the larger value, and the lifestyle of the subject is recorded in that order.

Node Y ₁₄ (A) and node Y ₁₄ (D) are markers that substitute for the disease state and are responsible. The node Y ₁₃ (A) and the node Y ₁₃ (C) work for guiding the node Y ₁₄ (A), and the node Y ₁₃ (E) works for guiding the node Y ₁₄ (D). Similarly, the nodes Y ₁₂ (C), Y ₁₂ (A), and Y ₁₂ (D) work to guide the nodes Y ₁₃ (A), Y ₁₃ (C), and Y ₁₃ (E), respectively. Furthermore, the nodes Y ₁₂ (A), Y ₁₂ (C), and Y ₁₂ (D) are guided by the nodes Y ₁₁ (A), Y ₁₁ (B), and Y ₁₁ (D), respectively.

The history of the cell memory of the subject in FIG. 11 can be shown as the matrix of FIG. 12 when the case where the cell memory is introduced in each constant region is represented by + and the case where it is not introduced is represented by −. + And − in FIG. 12 correspond to the states of the nodes Y ₁₁ (A) to Y ₁₄ (E) in FIG. 11, respectively.

  Further, it is assumed that five types of related molecules G (A) to G (E) are detected as gene information of genes related to blood molecules A to E. From the results of the genome analysis, when n types of diversity exist in each gene in the compared population, the population is n! It can be classified into small groups (groups shown in the left frame in FIG. 11) as × 5 (! Means factorial).

  When performing a search for similar histories, the data is first classified into small groups according to the genome information, and then the approximate data is selected based on the homology of the matrix shown in FIG. The highest degree of homology is when all items in the matrix match. In the case of partial approximation, the approximations are ranked according to their magnitude. Since the record of lifestyle is pasted in the data of the matrix, the subject can select future lifestyle by referring to the lifestyle information of other similar people.

  In step S112, the estimation control unit 120 estimates and controls the state change. That is, future changes in the state of the living body of the subject are estimated, and intervention methods including preventive methods are proposed.

  The simplest proposal is to present the search result of step S111 to the target person as it is. That is, in the search process of step S111, the history of another person similar to the history of the target person is searched. If the searched other person has any disease or illness, there is a high probability that the subject will also have the same disease or illness. Therefore, the subject can prevent the same disease or illness from being caused by making the environmental condition different from the others by changing the lifestyle, for example.

  Of course, it is possible to go even further and actively build and provide future options for maintaining the health of the subject. For example, among environmental conditions such as stress, dietary habits, exercise habits, infection history, and treatment history, an element having a stronger causal relationship can be identified and proposed to be changed. Further, when it becomes possible to change the factors related to the gene, it is possible to propose the change.

  As described above, by retrieving those information of the living body other than the subject, which is similar to the cell memory information, the environmental information, and the genetic information of the subject of health care, it is appropriate for the subject to maintain the health. It is possible to reliably obtain such information.

  In the present specification, the steps describing the processes of the flowcharts are not limited to the processes performed in time series according to the order, but are not necessarily performed in time series, and are performed in parallel or individually. It also includes.

  The embodiments of the present technology are not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the present technology. For example, it is possible to combine two or more arbitrary embodiments.

  1 biometric information processing device, 11 selection unit, 12 determination unit, 13 measurement unit, 14 creation unit, 15 division unit, 16 identification unit, 17 estimation control unit, 51 biometric information processing device, 61 selection unit, 62 measurement unit, 63 Determining part, 64 division part, 65 identification part, 66 estimation control part

Claims (28)

Lifestyle information acquisition step of acquiring information on the lifestyle of the target person,
A history information acquisition step of acquiring information on the history of genetic modification or epigenetics modification of the living body of the subject,
A biometric information processing method comprising: a similarity analysis step of analyzing the acquired information regarding the lifestyle and the information regarding the history of the genetics modification or the epigenetics modification with a person other than the subject. The biometric information processing method according to claim 1, wherein the information regarding the lifestyle is information regarding a dietary habit, an exercise habit, or a stress state. Further comprising an acquisition step of acquiring information regarding a history of infection, treatment or medication of said subject,
In the similarity analysis step, the acquired information about the history of infection, treatment or medication, the information about the lifestyle, and the information about the history of the genetics modification or epigenetics modification are similar to those other than the subject. The biometric information processing method according to claim 1 or 2, wherein the sex is analyzed. Further comprising a genetic information acquisition step of acquiring the genetic information of the subject,
In the similarity analysis step, the acquired gene information, the information about the lifestyle, and the information about the history of the genetics modification or epigenetics modification are analyzed for the similarity to those other than the subject. Alternatively, the biological information processing method according to claim 2. The biological information processing method according to claim 4, wherein the genetic information is single nucleotide polymorphism and structural polymorphism. The biological information processing method according to any one of claims 1 to 5, wherein the information regarding the history of the genetic modification or the epigenetics modification is acquired based on a change over time in the expression level of the biomolecule. The biological information processing method according to claim 6, wherein the biomolecule is a blood molecule. The information regarding the history of the genetic modification or the epigenetic modification is acquired based on the change over time in the blood glucose level measured by a glucose meter. Biometric information processing method. The biological information processing method according to any one of claims 1 to 8, wherein in the similarity analysis step, a temporal change of a biological state is discretized from information regarding a history of the genetics modification or the epigenetics modification. . The biological information processing method according to claim 9, wherein in the similarity analysis step, a baseline change in the biological state is estimated from a temporal change in the discretized biological state. The biological information processing method according to claim 10, wherein the baseline change is an irreversible change. The biometric information processing method according to claim 10 or 11, wherein the baseline change in the biological state is a change from a healthy state to a disease state. The biological information according to any one of claims 1 to 12, wherein in the similarity analysis step, a correlation between the information on the lifestyle and the information on the history of the genetics modification or epigenetics modification is analyzed. Processing method. The biological information processing method according to claim 13, wherein in the similarity analysis step, the correlation is analyzed by a clustering method. In the similarity analysis step, information of a plurality of individuals is recorded by classifying it into any of a group of information regarding the lifestyle, the genetic information, and information regarding the history of genetics modification or epigenetics modification. The biological information processing method according to claim 4, wherein the biological information processing is performed by referring to a database. The biometric information processing method according to claim 15, wherein the group of information on lifestyle habits is a group divided based on a diet habit, an exercise habit, or a stress state. The biological information processing method according to claim 15, wherein the group of genetic information is a group divided based on genetic diversity. The biological information processing method according to claim 15, wherein the group of information regarding the history of genetics modification or epigenetics modification is a group that is divided based on a change in a biological state. The biometric information processing method according to any one of claims 15 to 18, wherein the information of the plurality of individuals is recorded in the database in a format in which the privacy of each individual is concealed. The biometric information processing method according to claim 1, further comprising an output step of outputting a result of the similarity analysis. The biological information processing method according to claim 20, wherein in the output step, information about a disease or a disease other than the target subject that is similar and is identified based on the result of the similarity analysis is output. The biological information processing method according to any one of claims 1 to 21, wherein a future biological state of the subject is estimated based on a result of the similarity analysis. The biological information processing method according to any one of claims 1 to 22, wherein a disease or disease risk of the subject is identified based on a result of the similarity analysis. The biological information processing method according to claim 23, which outputs information regarding treatment or prevention of the disease identified based on the result of the similarity analysis. 25. The biometric information processing method according to claim 1, wherein a lifestyle improvement proposal is output for a disease or disease risk of the subject identified based on the result of the similarity analysis. . The biological information processing method according to any one of claims 1 to 25, wherein the genetic modification comprises DNA mutation or structural change, and the epigenetic modification comprises DNA methylation or histone modification. A lifestyle information acquisition unit that acquires information about the lifestyle of the target person,
A history information acquisition unit that acquires information regarding the history of genetic modification or epigenetics modification of the living body of the subject,
A biometric information processing apparatus, comprising: a similarity analysis unit that analyzes the acquired information regarding the lifestyle and the information regarding the history of the genetics modification or the epigenetic modification with a person other than the subject. Lifestyle information acquisition step of acquiring information on the lifestyle of the target person,
A history information acquisition step of acquiring information on the history of genetic modification or epigenetics modification of the living body of the subject,
A program for causing a computer to execute a process including a similarity analysis step of analyzing the acquired information regarding the lifestyle and the information regarding the history of the genetics modification or the epigenetics modification with a person other than the subject. .

Images