EA001095B1 - Computational method for designing chemical structures having common functional characteristics - Google Patents

Abstract

1. A computer-based method of designing chemical structures having a preselected functional characteristic, comprising the steps of: (a) producing a physical model of a simulated receptor phenotype encoded in a linear charater sequence, and providing a set of target molecules sharing at least one quantifiable functional characteristic; (b) for each target molecule; (i) calculating an affinity between the receptor and the target molecule in each of a plurality of orientations using an effective affinity calculation; (ii) calculating a sum affinity by summing the calculated affinities; (iii) identifying a maximal affinity; (c) using the calculated sum and maximal affinities to: (i) calculate a maximal affinity correlation coefficient between the maximal affinities and the quantifiable functional characteristic; (ii) calculate a sum affinity correlation coefficient between the sum affinities and the quantifiable functional characteristic; (d) using the maximal correlation coefficient and sum correlation coefficient to calculate a fitness coefficient; (e) altering the structure of the receptor and repeating steps (b) through (d) until a population of receptors having a preselected fits coefficient are obtained; (f) providing a physical model of a chemical structure encoded in a molecular linear character sequence, calculating an affinity between the chemical structure and each receptor in a plurality of orientations using said effective affinity calculation, using the calculated affinities to calculate an affinity fitness score; (g) altering the chemical structure to produce a variant of the chemical structure and repeating step (f); and (h) retaining and further altering those variants of the chemical structure whose affinity score approaches a preselected affinity score. 2. The method according to claim 1 wherein the step of producing a simulated receptor genotype comprises generating a receptor linear character sequence which codes for spatial occupancy and charge, and wherein the step of producing a physical model of a chemical structure comprises generating said molecular linear character sequence which codes for spatial occupancy and charge. 3. The method according to claim 2 wherein said effective affinity calculation comprises two measures, the first being a proximity measure wherein the proportion of uncharged portions on said simulated receptors being sufficiently close to non-polar regions on said molecular structure to generate effective London dispersion forces is estimated, and the second being the summed strengths of charge-dipole electrostatic force interactions generated between charged portions of said simulated receptor and dipoles present in said molecular structure. 4. The method according to claim 2 wherein said step of calculating the affinity fitness score includes calculating a sum and maximal affinity between the molecular structure and each receptor, the fitness score being calculated as: Σ {(|calculated maximal affinity - target maximal affinity|/ target maximal affinity} and wherein said preselected fitness score is substantially zero. 5. The method according to claim 2 wherein said step of calculating the affinity fitness score includes calculating a sum and maximal affinity between the molecular structure and each receptor, the fitness score being calculated as: Σ {(|calculated maximal affinity-target maximal affinity|/ 2 x target maximal affinity) + (|calculated sum affinity target sum affinityl/2 x target sum affinity|)}, and wherein said preselected fitness score is substantially zero. 6. The method according to claim 2 wherein said sum affinity correlation coefficient is rSA<2>, said maximal affinity correlation coefficient is rMA<2> , and wherein said fitness coefficient is F=(rMA<2> x rSA<2>)<0.5>, and wherein said preselected fitness coefficient is substantially unity. 7. The method according to claim 2 wherein said sum affinity correlation coefficient is rSA-MA<2>, said maximal affinity correlation coefficient is rMA<2>, and wherein said fitness coefficient is F=(rMA<2> x (1-rSA-MA<2>))<0.5> and wherein said preselected fitness coefficient is substantially unity. 8. The method according to claim 2 wherein said molecular linear character sequences comprise a plurality of sequential character triplets, a first character of said triplet being randomly selected from a first character set specifying position and identity of an occupying atom in a molecular skeleton of said molecular structure, a second character of said triplet being randomly selected from a second character set specifying the identity of a substituent group attached to said occupying atom, and a third character of said triplet being randomly selected from a third character set specifying the location of said substituent on the atom specified by said first character of the triplet. 9. The method according to claim 8 wherein the molecular linear character sequence is decoded using an effective molecular assembly algorithm which sequentially translates each triplet from said molecular linear sequence and thereafter fills unfilled positions on said molecular skeleton with hydrogen atoms. 10. The method according to claim 9 wherein the step of mutating said molecular structure includes at least one of the following steps: i) mutating said molecular genotype by randomly interchanging at least one of said first, second and third characters of at least one triplet from the associated character sets, ii) deletion wherein a triplet from molecular genotype is deleted, iii) duplication wherein a triplet in the molecular genotype is duplicated, iv) inversion wherein the sequential order of one or more triplets in the molecular genotype is reversed, and v) insertion wherein a triplet from the molecular genotype is inserted at a different position in the molecular genotype. 11. The method according to claim 10 wherein the step of mutating said molecular genotypes includes recombining randomly selected pairs of said retained mutated molecular genotypes whereby corresponding characters in said molecular linear sequences are interchanged. 12. The method according to claim 2 wherein each character in the receptor linear character sequence specifies one of either a spatial turning instruction and a charged site with no turn. 13. The method according to claim 12 wherein said receptor phenotype comprises at least one linear polymer provided with a plurality of subunits, one of said subunits being a first subunit in said at least one linear polymer. 14. The method according to claim 13 wherein said receptor linear character sequence is decoded using an effective receptor assembly algorithm in which turning instructions applied to each subunit subsequent to said first subunit are made relative to an initial position of said first subunit. 15. The method according to claim 14 wherein said characters specifying spatial turning instructions code for no turn, right turn, left turn, up turn, down turn, and wherein characters specifying charge sites code for positively charged site with no turn, and negatively charged site with no turn. 16. The method according to claim 14 wherein said subunits are substantially spherical having a Van der Waals radii substantially equal to the Van der Waals radius of hydrogen. 17. The method according to claim 15 wherein the step of mutating said receptor genotype includes at least one of the following steps: i) deletion wherein a character from the receptor genotype is deleted, ii) duplication wherein a character in the receptor genotype is duplicated, iii) inversion wherein the sequential order of one or more characters in the receptor genotype is reversed, and iv) insertion wherein a character from the receptor genotype is inserted at a different position in the genotype. 18. The method according to claim 17 wherein the step of mutating said receptor genotypes includes recombining randomly selected pairs of said retained mutated receptor genotypes whereby corresponding characters in said receptor linear sequences are interchanged. 19. A method of screening chemical structures for preselected functional characteristics, comprising: a) producing a simulated receptor genotype by generating a receptor linear character sequence which codes for spatial occupancy and charge; b) decoding the genotype to produce a receptor phenotype, providing at least one target molecule exhibiting a selected functional characteristic, calculating an affinity between the receptor and each target molecule in a plurality of orientations using an effective affinity calculation, calculating a sum and maximal affinity between each target molecule and receptor, calculating a sum affinity correlation coefficient for sum affinity versus said functional characteristic of the target molecule and a maximal affinity correlation coefficient for maximal affinity versus said functional characteristic, and calculating a fitness coefficient dependent on said sum and maximal affinity correlation coefficients; c) mutating the receptor genotype and repeating step b) and retaining and mutating those receptors exhibiting increased fitness coefficients until a population of receptors with preselected fitness coefficients are obtained; thereafter d) calculating an affinity between a chemical structure being screened and each receptor in a plurality of orientations using said effective affinity calculation, calculating an affinity fitness score which includes calculating a sum and maximal affinity between the compound and each receptor and comparing at least one of said sum and maximal affinity to the sum and maximal affinities between said at least one target and said population of receptors whereby said comparison is indicative of the level of functional activity of said chemical structure relative to said at least one target molecule. 20. The method according to claim 19 wherein said effective affinity calculation comprises two measures,

Description

This invention relates to methods involving the creation of chemical structures having common useful functional properties using a computer and based on specific combinations of spatial configuration and binding affinity. More specifically, this invention provides a method for producing computer-modeled receptors that functionally mimic biological receptors. Simulated receptors are created in order to provide optimal selective affinity for known target molecules. Then, chemical structures are also generated that exhibit selective affinity for simulated receptors.

BACKGROUND OF THE INVENTION

Biological receptors are linear polymers of either amino acids or nucleotides, the chains of which are stacked with the creation of three-dimensional membranes for binding the substrate. The location of these specific three-dimensional linear structures and the placement of charged sites on the surface of the shells is the result of evolutionary choices based on functional efficiency.

The selectivity of biological receptors depends on the magnitude of the attractive and repulsive forces arising between the receptor and the substrate. The magnitude of these forces varies with the magnitude and proximity of the charged sites on the surfaces of the receptor and substrate.

Since substrates differ in the number and size of charged sites that are present or induced on these surfaces, as well as in the spatial arrangement of these sites, binding affinity may vary depending on the structure of the substrate. Substrates with similar affinity for binding to the same receptor are very similar to each other in the spatial arrangement of at least some induced and fixed charged sites. If receptor function is correlated with binding affinity, then substrates with similar binding affinity should be functionally similar in effect. In this sense, we can say that the receptor recognizes or quantifies the similarity of the substrates.

The traditional methods used for molecular recognition of identity or the determination of new chemical compounds or substrates with selective affinity for binding to receptors are based on finding molecular common subgraphs of active substrates and using them to predict new similar compounds. The disadvantage of this technique is that it is based on the assumption that substrates exhibiting similar binding efficiency are structurally similar. However, in many cases, substrates that differ in structure can exhibit a similar affinity for binding to the same receptor. Later methods based on quantification of structural activity ratios (0§AK) are suitable only for creating new compounds of one class of structures and are hardly suitable for creating new molecular structures exhibiting the desired selective affinity, see, for example, Beai, ΡΗίΙίρ M. , Mo1esi1ag KesodshNoey: Thie Meakiget Sha Zeat Roth Mo1esi1ag ShtjatNu ίη Idaii-Keker! From 1 and 1geasNoey, ίη Soicerkiy ArrNsaNoik oG Mo1esiag ZipPagNu. Her. Magk A. 1okikoi aiy Oeta 1st M. Maddyuta, r. 211-238 (1990).

Recent attempts have focused on creating atomic models or pseudoreceptors in which atoms and functional groups are linked, or mini-receptors, which are unconnected sets of atoms or functional groups (§iuyet, ΤΡ. (1993) Ιη 3Ό 0§AK ίη Aegis Be51dp: Thyot, Meyyuk aiy ArrBsaNoik; Kiysh, N.Ye .; Exot, BeIet Ρ. 336). Other methods include creating around the ligands a known target of a shell from a number of atoms of the model and calculating the intermolecular forces arising between the ligand and the receptor model. Such models are characterized by a high degree of correlation between the calculated binding energy and biological activity (^ a11egk, B.E. Aiy Nshyk, K.M. (1994) 1. Meyu. Sket. 37: 2527), but they have not been developed to such an extent when it is possible to obtain new chemical structures with selective affinity for receptor models.

Therefore, it would be very desirable to create a method for determining non-trivial similarities between different chemical structures, which is sufficient and necessary to explain their general properties, which can then be used as a basis for creating new chemical structures with useful functional properties, based on specific combinations of spatial configuration and binding affinities.

SUMMARY OF THE INVENTION

This invention provides a method for determining the non-trivial similarity of various chemical structures, which is necessary and sufficient to explain their identical functional properties. The method also involves the creation of new chemical structures that exhibit similar functional properties.

The principle underlying this invention is to use a two-stage computer method to create or discover chemical structures with useful functional properties based on specific combinations of steric configuration and binding affinity. At the first stage of this process, algorithmic emulation of antibody formation is used to create a population of computer-simulated receptors that mimic biological receptors with optimized binding affinity for selected target substrates. In the second stage of the process, simulated or virtual receptors are used in order to evaluate the affinity for binding of existing compounds or to create new substrates with optimal binding.

The method described in this application provides for the creation of simulated receptors that mimic selected features of biological receptors, including evolutionary processes that optimize their selectivity for binding. Mimetics or simulated receptors created in this way can be used to recognize specific similarities between molecules. Like antibodies and other biological receptors, the modeled receptors of the invention are feature extraction mechanisms: they can be used to identify or recognize common or similar structural features of target substrates. Binding between receptors and target substrates is used to recognize traits. Substrates can be quantified based on the amount of binding to a specific modeled receptor. Compounds having common specific structural features will also have similar binding to the same virtual receptor.

The affinity between biological receptors and substrates is determined by the steric correspondence between adjacent surfaces of the receptor and substrate, the absence of water between the nonpolar sites of the two surfaces and the magnitude of the electrostatic forces arising between adjacent charged sites. In some cases, the formation of covalent bonds between the substrate and the receptor also contributes to the amount of binding. Modeled receptors obtained in this way mimic the binding mechanisms of their biological counterparts. The average proximity of the surfaces of the receptor and the target and the magnitude of the electrostatic attraction force arising between the charged centers on both surfaces are used to determine the binding affinity. The obtained values are used to assess the molecular similarity of the substrates.

The amount of binding may generally depend on the interaction between the entire surface of the substrate and the closed receptor or receptor membrane completely surrounding the substrate. In this case, an analysis of the general similarity between the substrates is the basis for establishing useful quantitative structure-activity relationships. However, in most, if not all biological systems, affinity is determined locally rather than globally. In general, the interaction between substrate molecules and biological receptors is limited to contacts between isolated fragments of the surfaces of the receptor and substrate. In this situation, the study of general (global) similarity between substrates is not suitable as a method for studying the structure-activity relationship, since only fragments of the substrate participate in the formation of binding.

Structures with local similarity have the same structural fragments in the same positions and orientations. Structures having local affinity do not necessarily have global affinity. A sample of molecular properties can be achieved by a complete sampling method, including an assessment of global similarity; the method of fragmented sampling, including the assessment of local similarity, and the method of fragmented multiple sampling, including the assessment of both local and global similarity.

The analysis of local similarity is based on sampling of discrete sections of substrates for similar structures and determining the distribution of charges. In biological receptors, local sampling is possible due to irregularities or irregularities in the adjacent surfaces of the substrate and receptor. Interactions between closely spaced opposing surfaces will prevail over interactions between more distant sites in determining binding affinity. The proximity of adjacent surfaces will also determine the strength of hydrophobic binding. Effective simulated receptors obtained by this method should use local discrete sampling of target substrates (molecules) to assess the functionally relevant similarity of the compounds.

The determination of local similarity is complicated by two factors: 1) the number, location and identity of the relevant fragments, necessary and sufficient for specific binding, usually cannot be established by simple deduction based on the chemical structure of the substrate; and 2) the position and orientation of the fragments selected during the analysis depend on the structure of the whole molecule.

Part of this method, aimed at obtaining simulated receptors that can classify the similarity between chemical substrates, is mainly a search for receptors that are samples of relevant fragments of substrates in a relevant position in space. The optimization process is based on four features of simulated receptors: 1) universality: the ability of receptors to bind more than one substrate; 2) specificity: receptor binding depends on the structure of the substrate; 3) cost-effectiveness: receptors distinguish substrates by the minimum set of local structural features; and 4) mutability: a change in the structure of the receptor can alter its affinity for binding to a specific substrate. Encoding the receptor phenotype in the form of a linear genotype represented by a characteristic strand facilitates the processes of mutation, recombination, and inheritance of the structural characteristics of simulated receptors.

Simulated receptors that satisfy these basic criteria can be optimized to achieve specific binding to locally similar substrates using evolutionary selective reproduction techniques. This is done by encoding the spatial configuration and distribution of the charged receptor sites in an inherited format that can undergo changes or mutations. Like biological receptors, simulated receptors obtained by this method define a three-dimensional alienation space. Such a three-dimensional space can be outlined to an arbitrary degree of resolution by a one-dimensional line of sufficient length and tortuosity. Proteins formed by linear polymers of amino acids are examples of such structures. Similarly, the three-dimensional structure of simulated receptors can be encoded as a linear formation of changing teams. This one-dimensional encoded form of the receptor is its genotype. The decoded form used to achieve binding is its phenotype. In the process of optimization, changes (mutations) of the receptor genotype are made. The effect of these changes on the affinity of the phenotype for binding is then evaluated. Genotypes generating phenotypes with a given binding affinity are saved for further change until the desired degree of phenotype optimization is achieved through iteration of the mutation and selection process. A variety of evolutionary methods can be used to create simulated receptor populations with optimal binding values, including classical genetic receptor algorithms obtained by this method, then used to generate or identify new chemical structures (compounds) that have specific useful properties of molecular targets used as selection criteria for obtaining simulated receptors. When using interactions with receptors as a selection criterion, new chemical structures are created that are optimal for the receptors. Since these structures must satisfy the requirements necessary and sufficient to ensure selectivity of the receptors, they must also have biological activity similar to the activity of the original molecular targets. A population of simulated receptors with improved selectivity can also be used to screen existing chemical structures for compounds with high affinity, which may have these beneficial properties. The same method can be used to screen compounds with selected toxicological or immunological properties.

According to one aspect of the present invention, there is provided a computer method for creating chemical structures having predetermined functional characteristics, including operations:

(a) creating a physical model of the phenotype of a simulated receptor encoded in a linear characteristic sequence and creating a set of target molecules having at least one common quantitative functional characteristic;

(b) for each target molecule:

(ί) determining the affinity between the receptor and the target molecule in each of the multiple orientations using the calculation of the effective affinity;

(ίί) determining the total affinity by summing the calculated affinity values;

(w) identification of the maximum affinity;

(c) using the calculated total and maximum affinity values for:

(ί) determining the maximum affinity correlation coefficient between the maximum affinity values and the quantitative functional characteristic;

(ίί) determining the total affinity correlation coefficient between the total affinity and the quantitative functional characteristic;

(d) using the maximum correlation coefficient and the total correlation coefficient to determine the correspondence coefficient;

(e) changes in the structure of the receptor and the repetition of operations (b) - (d) until a population of receptors with a given correspondence coefficient is obtained;

(f) creating a physical model of the chemical structure encoded in a molecular linear characteristic sequence, determining the affinity between the chemical structure and each receptor in a variety of orientations using the specified definition of effective affinity, using certain affinity values to determine the degree of affinity correspondence;

(g) changing the chemical structure to obtain a variant of the chemical structure and repeating operation (e); and (h) preservation and further change of those variants of the chemical structure, the degree of affinity of which is approaching a given value.

According to another aspect of the invention, there is provided a method for screening chemical structures to obtain desired functional properties, including:

(a) obtaining a genotype of a simulated receptor by generating a linear sequence of receptor traits that encodes the spatial arrangement and charge;

(b) decoding the genotype to obtain the receptor phenotype, creating at least one target molecule exhibiting a given functional characteristic, calculating the affinity between the receptor and each target molecule in a variety of orientations using the calculation of effective affinity, finding the total and maximum affinity between each molecule target and receptor, determination of the total affinity correlation coefficient for the dependence of the total affinity on the specified functional characteristics of the target molecule and and the maximum affinity correlation coefficient for the dependence of the maximum affinity on the specified functional characteristics and the calculation of the compliance coefficient depending on the specified total and maximum affinity correlation coefficients;

(c) mutation of the genotype and repetition of the operation (b) and preservation and mutation of those receptors that have increased correspondence coefficients until a population of receptors with predetermined matching coefficients is formed;

(d) calculating the affinity between the chemical structure being screened and each receptor in a variety of orientations using the indicated effective affinity calculation, calculating the degree of affinity correspondence, which includes calculating the total and maximum affinity between the compound and each receptor and comparing at least one of the indicated total and maximum affinity with the total and maximum affinity of at least one target and said population of receptors, wherein said comparison S THE indicator of the level of functional activity of said chemical structure with respect to at least one target molecule.

According to another aspect of the invention, there is provided a method for creating simulated receptors that mimic biological receptors with selective affinity for compounds with similar functional characteristics, including operations:

(a) obtaining the genotype of a simulated receptor by generating a linear receptor sequence that encodes a spatial arrangement and charge;

(b) decoding the genotype to obtain the receptor phenotype, creating a set of target molecules with similar functional characteristics, calculating the affinity between the receptor and each target molecule in a variety of orientations using an effective affinity calculation, calculating the total and maximum affinity between each target molecule and receptor, calculating the total affinity correlation coefficient for the dependence of the total affinity on the functional characteristic for each target molecule and the maximum correlation coefficient for the dependence of the maximum affinity from the affinities of said functional characteristics for each of the target molecule and calculation of the corresponding coefficients in dependence on said sum and maximal affinity correlation coefficients for each of the target molecule; and (c) mutations of the genotype and repetition of stage (b) and the conservation and mutation of those receptors that have increased correspondence coefficients until a receptor population with predetermined matching coefficients is obtained.

According to a further aspect of the present invention, there is provided a computer method for creating chemical structures with a given functional characteristic, including operations:

(a) creating a physical model of the receptor and a set of target molecules, the latter having at least one quantifiable functional characteristic;

(b) for each target molecule:

(ί) determining the affinity between the receptor and the target molecule in each of the multiple orientations using the calculation of the effective affinity;

(ίί) calculating the total affinity by summing the found affinity values;

(ίίί) determining maximum affinity;

(c) using calculated total and maximum affinity values in order to:

(ί) calculate the maximum affinity correlation coefficient between the maximum affinity and the quantifiable functional characteristic;

(c) calculate the total affinity correlation coefficient between the total affinity and the quantifiable functional characteristic;

(d) using the maximum correlation coefficient and the total correlation coefficient to calculate the correspondence coefficient;

(e) changes in the structure of the receptor and repetition of operations (b) - (d) until a population of receptors is formed with a given correspondence coefficient;

(f) creating a physical model of the chemical structure, determining the affinity between the chemical structure and each receptor in a variety of orientations using the indicated calculation of effective affinity, using the calculated affinity values to find the degree of affinity correspondence;

(g) changing the chemical structure to create a variant of the chemical structure and repeating stage (e); and (h) the preservation and further change of those variants of the chemical structure whose degree of affinity correspondence approaches a predetermined value.

According to yet another aspect of the invention, there is provided a method for encoding chemical structures comprising atomic elements, the method comprising creating a linear characteristic sequence that encodes the spatial arrangement and charge for each atom of said chemical structure.

Brief Description of the Figures

Below the invention will be illustrated by example with reference to the accompanying figures.

In FIG. 1 is a diagram showing the relationship between creating a genotype code and translating to obtain the corresponding phenotype, reflecting part of the present invention;

in FIG. 2 shows a general diagram illustrating receptor optimization operations for selectively binding to a number of substrates using point mutations, reflecting part of the present invention;

in FIG. Figure 3 shows a general flow diagram of a method for producing a population of receptors with optimal affinity for selective binding to a number of chemical substrates and using these optimized receptors to produce a number of new chemical substrates with common identical functional characteristics;

in FIG. 4a shows several chemical compounds used in the example related to ligand generation;

in FIG. 4b shows ligands 1.1-1.4 obtained by the method of this invention in an example illustrating ligand generation, where each ligand has at least one orientation in which it is structurally similar to benzaldehyde;

in FIG. 4c shows ligands 2.1-2.4 obtained by the method of this invention in an example illustrating the generation of ligands related to the creation of chemical structures that are effective in repelling mosquitoes.

Description of Preferred Options

This method can be divided into two parts: (A) the evolution of a population of simulated receptors with selective affinity for compounds with common functional properties; and (B) the formation of new chemical structures having common functional properties. Part (A) consists of several operations, including 1) generation of the receptor genotype and phenotype; 2) target generation; 3) the presentation of known chemical structures (s) to the receptor; 4) an assessment of the affinity of the receptor for the chemical structure (s); 5) an assessment of the selectivity of the receptor with respect to the chemical structure (s); 6) the evolution of a family of related receptors with optimized selective affinity for the chemical structure (s); screening of chemical substrates to assess toxicology and pharmacological activity and the use of optimized receptors to create a new chemical structure (s) with selective affinity for receptors.

The following description of a better embodiment of the invention relates to various tables of molecular and atomic radii, polarizabilities, effective dipole values, and transition and coupling factors, these values are given in table. 1-5 are given at the end of the description. Schemes illustrating calculation examples, not limiting the invention, are presented at the end of the description in the form of modules 1 -1 4.

Part A: Evolution of a population of simulated receptors with selective affinity for target molecules having common functional properties (1) Generation of the receptor genotype and phenotype code

Both genotypes and phenotype of the simulated receptor are calculated objects. The phenotypes of simulated receptors consist of folded, unbranched polymers of spherical subunits, the diameter of which is equal to the van der Waals radius for atomic hydrogen (-110 pm). Subunits can be connected to each other at any two of six points corresponding to segments of spheres with each of their main axes. In this case, the bonds between the subunits cannot stretch or rotate, and the centers of two connected subunits are always separated by a distance equal to the length of their sides (i.e., radius 1 of hydrogen). When two subunits do not adhere to opposite surfaces of their common neighbor, turns (turns) occur. Four types of orthogonal turns are possible: left, right, upper and lower. Turns should be parallel to one of the main axes. To simplify the calculations, when the turns intersect with other subunits in the polymer, it is assumed that the subunits occupy the same space with other subunits.

A simulated receptor consists of one or more discrete polymers. In the case of receptors consisting of many polymers, individual polymers can be located at different points in space. To simplify, in this case, it is assumed that all polymers constituting the same receptor have the same length (= the number of subunits). This limitation is not a requirement for functionality, and polymers of different lengths may be suitable for modeling specific systems.

The structure of each polymer is encoded as a sequential set of changing instructions. These commands define individual changes with respect to the internal designation frame based on the initial orientation of the first subunit in each polymer.

In this case, the receptor and substrate hydrates are not processed; instead, it is assumed that any water molecules present at the binding site are constantly attached to the surface of the receptor and enter its structure. This is an arbitrary approximation, and it will be apparent to those skilled in the art that a more accurate processing can be carried out (see, for example, Alcoz. 1995, Moecius1agentipod 32: 199211).

As can be seen from FIG. 1, the code generation module generates arbitrary feature threads. Each sign represents a changing team or determines the characteristics of the charge or reactivity at a point in three-dimensional form, which is a virtual receptor. To create a thread that describes the three-dimensional shape of the receptor in the Cartesian (rectangular) coordinate system, at least five different attributes are required. Other systems, such as tetrahedral structures, can also be created using various sets of variable commands. Attributes represent changing teams that are defined with respect to the current direction in the structure of the virtual receptor in three-dimensional space (i.e., the commands refer to the notation frame inherent in the virtual receptor, and not to an arbitrary external notation frame).

Changing commands are given with respect to the direction of the current and the orientation of the polymer. Only left, right, top and bottom turns are allowed. If the rotation does not occur, the polymer chain either ends or continues in the same direction.

For a rectangular system, there is the following minimum set of features: C ₁ = no rotation; C ₂ = right turn; C ₃ = left turn; C ₄ = upper turn and C ₅ = lower turn. It is clear that teams can be combined to create diagonal turns, for example, L _{1> 2} = 0С ₂ ; L _{2> 1} = C ₂ C, etc. The number of different features that define different charges or reactive states is not limited and can be selected in accordance with empirical data. Codes may differ from each other in length (number of features) and the frequency with which specific features appear in series.

An example of creating a genotype.

It will be apparent to those skilled in the art that the following example of creating a genotype code and expressing a phenotype is only illustrative. In this example, the following conditions apply.

(1) The character set used to generate the codes consists of five characters related to changing teams and two characters identifying a charged site: 0 = no rotation; 1 = right turn; 2 = top turn; 3 = left turn; 4 = bottom turn; 5 = positively charged site (no turnaround) and 6 = negatively charged site (no turnaround).

(2) Subunits are divided into two types: charged and uncharged. It is assumed that all charged subunits carry a single positive or negative charge. A uniform charge is an arbitrary condition.

(3) Receptors contain 15 discrete polymers. The length of the complete code is always a multiple of fifteen. The length of each polymer is equal to the length of the entire code divided by fifteen. Obviously, receptors can be created from any number of discrete polymers with a variable or constant length.

(4) The following parameters are defined by the user: (a) total code length (and polymer length); (b) the frequency with which each character appears in the code; and (c) combinations of characters. Module 1 provides an example of a scheme for creating a genotype code.

An example of creating a receptor phenotype.

Each genotype code is translated to create a three-dimensional description of the corresponding phenotype or virtual receptor. To convert the changing commands into a series of coordinate triplets that describe the position of consecutive subunits representing receptor polymers, a translation algorithm from a given starting point is used. The initial coordinates for each polymer must be set prior to translation. Translation assumes that the centers of consecutive subunits are separated from each other by a distance equal to the covalent diameter of the hydrogen atom.

The translation algorithm reads the coded thread sequentially with the formation of sequential turns and straight segments. The interpretation of successive rotations with respect to the external coordinate system depends on the previous sequence of rotations. It is assumed that for each polymer forming the receptor, the initial orientation is the same. In this case, the translation algorithm is described in table. 1, showing input and output states. If no rotation is formed, the latest values for Ax, Au, Δζ and new states are used to determine the new coordinate triplet. Charge sites are processed as straight (without turns) segments. The initial value of the old state is 20.

The user can define the following parameters:

but. The initial coordinates for each polymer forming the receptor.

The output is stored as:

a) three vectors (one for each axis {x _1, x2, Xs, x}, {y1, Vn}, {Ζ1, ..., ζη }).

b) three-dimensional binary matrix.

c) separate vectors for the coordinates of charge sites. An example of code translation is given in module 2.

(2) Target Generation

Targets are presented in the form of molecules consisting of spherical atoms. It is believed that atoms are solid spheres with fixed radii for each type of atom. The radius of the solid sphere, at which the repulsive force acting between the target atoms and the virtual receptor, is considered to be an infinitely large value, is approximated by the Van der Waals radius given in Table.

2. Other approximate values of the radius of van der Waals can be used instead of the values given in table. 2.

The distance between the atomic centers of two atoms linked by a covalent bond is expressed as the sum of the radii of their covalent bonds. The radii of covalent bonds change with a change in the order of the bonds in the form of atoms. Examples of suitable radii are given in table. 3. As a first approximation, it is assumed that the bond length is fixed (ie, vibration of the bonds is ignored). Rotation of bonds is allowed, and multiple configurations of the same structure are required to select representative rotational states. Configuration stability is not considered, since binding to a virtual receptor can stabilize energetically unstable configurations. For the formation of target ligands, various energy minimization algorithms can be used.

Electric charges due to dipole moments of bonds are considered localized in atomic nuclei. A negative charge is carried by an atom with a greater magnitude of electronegativity. The dipole values used in this case are given in table. 4. Instead of the values indicated in the table. 4, other dipole values may be used.

(3) Presentation of targets

The affinity of each target for the simulated receptor (s) is tested for several target orientations relative to the upper surface of the receptor. The upper surface is determined by the shift algorithm. Before evaluating the binding affinity, the target and receptor must be brought into contact. Contact occurs when the distance between the centers of at least one receptor subunit and at least one target atom is equal to their combined van der Waals radii. In order to determine the relative positions of the target and the receptor at the contact point, the target is moved incrementally towards the surface of the receptor along a path perpendicular to the surface and passing through the geometric centers of both the receptor and the target. When contact occurs, the target reaches a position of collision with the receptor. The moved target atoms, when the collision position is reached, are used to calculate the distances between the target atoms and the receptor subunits. These distances are used to determine the strength of electrostatic interactions and proximality.

In this case, it is assumed that the target moves in a straight line towards the receptor and retains its initial orientation at the moment of contact. An alternative approach would allow the target to change its orientation incrementally as it approaches the receptor so that the position of maximum affinity is reached at the point of contact. Although this method is functionally similar to that used, it is much more complicated to compute. In this case, multiple orientations are tested using a simpler computing device. This method provides for controlled displacement of the trajectory along the x axis and / or at the receptor to accommodate larger molecules. This is required to increase selectivity when molecules of different sizes are tested on the same receptor.

Before determining the contact position, the target orientation is randomized by random rotation with an increment of 6 ° around each of the x, y, and ζ axes. You can use larger or smaller increment values. In this test series, each of these random target orientations is unique. The reliability of the optimization process depends on the number of orientations of the target used, as well as on the number of estimated target compounds. An example implementation of the method of target presentation is shown in module 3.

(4) Calculation of affinity

Approximation Strategy.

This embodiment of the method is based on a simplified approximation, which boils down to assessing the main components involved in the binding using a relatively simple computing device. Approximation is described in the following sections. However, it will be apparent to those skilled in the art that more accurate calculation methods can be used that provide a more accurate affinity value. Known software packages for determining more accurate affinity values can be used in this method.

Studies of crown ethers show that the electron density distribution of small molecules can be used to describe the electron densities of compounds with large molecules (Vtishpd, N. GPB Rey, Ό. (1991) 1. Sosti !. Seyes. 12: 1). To determine the distribution of strictly local charges, characterized by a charge and a dipole moment, the method can be used (NyzGeM, R.L. (1977) Tjeot. Sjes. As! A 44: 129). The result is a distribution of the total electron density of the molecule into overlapping atomic parts whose dimensions are related to the radii of free atoms.

On the example of crown ethers, it can be shown that the main components of the electrostatic interaction are determined more by local rather than global movement of charges between atoms. The charge distribution is mainly distributed by the effect of small intervals due to various chemical bonds. In particular, non-neighboring atoms make a small contribution to the dipole moments of atoms. In addition, although the transfer of charges between atoms is also affected by the electrostatic field of the whole molecule, calculations for crown ethers show only a very small effect on the distribution of charges.

Certain quantities of atomic charges and dipole moments can be used to describe the electrostatic interaction (Vtishpd, N. GPB Rey, Ό. (1991) 1. Sotgi !. Set. 12: 1). In addition to the radius of van der Waals, only a small contribution is made by atomic quadrupole moments. The calculation of the electrostatic potential, which takes into account only atomic charges, gives very poor results, while the use of dipole moments makes it possible to obtain more accurate values.

Based on these considerations, we can say that the method according to this invention uses an approximation of the affinity between the target ligand and the simulated receptor (s), created on the basis of two values.

. The magnitude of the electrostatic forces acting between the charged subunits of the simulated receptor (s) and the atomic dipoles of the target ligand (chemical structure). It is assumed that since charged subunits transfer non-transferring unit charges, the magnitude of these forces is directly proportional to the magnitude of the atomic dipole and inversely proportional to the distance between the simulated receptor and the atomic dipole of the ligand.

2. The ratio of non-polar or uncharged subunits of the simulated receptor close enough to non-polar sections of the ligand for the formation of significant discrete forces in London.

Assumptions used to calculate (determine) affinity according to this embodiment of the method.

. It is assumed that the chemical target substrates studied according to this method are neutral (i.e. non-ionized) molecules. This is an arbitrary restriction, but the same technique can be used for charged and uncharged targets.

2. It is assumed that dipole moments are localized in atomic nuclei. A similar definition of affinity can be made by assuming that the dipole moment is in a covalent bond. According to AShdyat e! a1. (1989), these assumptions are functionally equivalent.

3. It is assumed that the environment surrounding the virtual receptor is a solvent in which the target is dissolved. The target is distributed between the solvent and the virtual receptor.

4. It is assumed that at the time of determining the affinity, the target and receptor are motionless in relation to each other and are in a specific fixed orientation.

5. It is assumed that targets interact only with two types of sites on the receptor surface: sites with fixed charges (positively or negatively charged) and non-polar sites.

Based on these assumptions, it is only necessary to consider the following factors that contribute to the magnitude of the interaction force.

1. A charge-dipole -O ² q ^2/6 (4ne) ^{2 4} CTG

2. Charge-nonpolar ip ² a / 2 (4ne) ² g ⁴

3. Dipole-non-polar (Debye energy) -ts ² o / (4pe) ² g ⁶

4. Non-polar-non-polar (energy of London) -.75 [yua ² / 4pe) ² g ⁶ ]

According to this method, only relative forces are considered in the approximation; therefore, all constants are neglected. In addition, it is assumed that the site with a fixed charge is unitary and either positively or negatively charged. Based on this, we can present the above expressions in the following simplified form:

1. Charge-dipole -ts ² / g ⁴ or -μ / g ²

2. Charge-non-polar -α / g ⁴

3. Dipole-non-polar (Debye energy) μ ² α / τ ⁶ or -μα ' ⁵ / Γ ³

4. Non-polar-non-polar (London energy) -a ² / g ⁶ or -α / g ³

In general, factors 2 and 3 make only a small contribution to the interaction. However, factors 1 and 4 significantly affect the magnitude of the interaction energy. According to this technique, it is assumed that most of the interactions between non-polar fragments occur between adjacent alkyl groups and the hydrogens of aromatic compounds and non-polar subunits of the receptor. It is assumed that under these conditions the value of α is constant.

Contribution of hydrophobic strength and exclusion of water.

When determining the affinity for binding, solvation effects are important. For example, the formation of a hydrophobic bond is based on close spatial association of nonpolar hydrophobic groups, while the contact between the hydrophobic sites and water molecules is reduced to a minimum. The contribution of the formation of a hydrophobic bond is equal to half the total strength of the antibody-antigen bonds. Hydration of the surfaces of the receptor and substrate is also a significant factor. Water bound to the polar sites on the surface of the receptor or substrate can affect binding or increase affinity due to the formation of transverse bridges between the surfaces.

Hydrophobic interaction characterizes a strong attraction between hydrophobic molecules in water. In the case of receptor-target interactions, the attraction between nonpolar fragments of the target and neighboring domains of non-polar receptor subunits should be noted. The effect appears mainly due to entropic factors leading to the rearrangement of surfaces with the exception of the presence of water between neighboring non-polar domains. There are no exact theoretical grounds for hydrophobic interaction; however, it has been established that the contribution of hydrophobic forces is up to 50% of the magnitude of the attractive force between antibodies and antigens. To determine the hydrophobic interaction between targets and virtual receptors, this technique involves the assessment of a portion of the receptor that is effectively protected from solvation by binding to the target. It is believed that all non-polar (uncharged) subunits that are no further than a fixed distance of non-polar target atoms are protected from solvation by solvent molecules whose diameter is equal to or greater than the limiting distance.

Calculation of total affinity.

The calculation of the total affinity used in this method combines two components of the interaction: the total charge-dipole interaction forces and the proximity value. It is assumed that in this case, both of these factors are isotropic. It will be apparent to those skilled in the art that a greater degree of resolution can be obtained if anisotropic affinity determinations are used, although the calculations are much more complicated.

The charge-dipole interaction is calculated as E = Tr | / g _|| ''. where μ ^ is the dipole moment of the ίth target atom and the distance between the ίth atom and _) is their charge site on the receptor, and the coefficient ν can be 2, 3, or 4. The contribution of Ό to the total affinity is more sensitive to charge separation for large quantities ν.

The proximation measure is defined as Ρ = Ση ₁ / Ν, where n ₁ = the number of uncharged receptor subunits, which are determined by the maximum distance d from the ί-th target atom with a dipole moment <0.75 Debye. In this method, q can vary from 1 to 4 diameters of the subunit (this value approaches the van der Waals radius of water). Ν indicates the total number of subunits forming the receptor.

The affinity value A is calculated from the values of Ό and P using the following equation Α = [Ρ (Ό + ΝΡ / Κ)] ^0.5 , where k denotes the correspondence constant (in this case, k = 1 0000). The value of P in the equation plays two roles. In the first case, this is a significant factor. As a measure of high degree of conformity, it is used to outweigh the magnitude of affinity for those configurations in which the non-polar portions of the target and receptor are in close contact. Under these conditions, the energies of hydrophobic interactions will be large and will significantly affect the stability and strength of the bond. Under these conditions, the target has fewer possible trajectories in order to avoid contact, and its retention time will be longer. In the second case, P is used to estimate the contribution of the dispersion energy to the interaction force. It is assumed that the dispersion energy will be significant only for uncharged, non-polar sites and that it will be significant only when the target and receptor are close to each other (i.e., when the distance between them is less than d). The quantities k and q can be selected so that it is possible to change the relative contribution of P and Ό. In general, P dominates in the case of nonpolar targets, and Ό is more significant for targets with large local dipoles. Hydrogen binding is approximated by paired negative and positively charged receptor units, which simultaneously interact with the hydroxyl, carboxyl or amine groups of the target.

Alternative approaches to determining the affinity-polarizability of a bond.

In some cases, it may be preferable to introduce into the affinity determination method a parameter corresponding to the relative polarizability of the target atoms. In this case, the equation for calculating P ₂ in the equality Α = [Ρ (Ό + ΝΡ ₂ / Ε)] ⁰ · ^{5 is} not: Ρ ₂ = Ση ₁ / Ν. Instead, P _{2 is} calculated by the equation P ₂ = Σα ₁ η ₁ / N, where n ₁ = the number of charged or uncharged receptor subunits, which are determined by the maximum distance d from the ίth atom of the target, and denotes the relative polarizability of the ίth atom of the target. For simplicity, a _H for aliphatic hydrogen is taken to be 1.0. The value of k must be selected if polarizability is used. The values of polarizabilities based on the sums of polarizabilities of neighboring bonds are given in table. 5.

Since the polarizability is related to the displacement of the electron cloud, the polarizability of a molecule can be calculated as the sum of the polarizabilities of its covalent bonds. This additivity is valid for non-aromatic molecules that do not have delocalized electrons.

An alternative technique is the specificity of functional groups.

The affinity approximation used in this case can be replaced by functionally similar calculations, in which the relation between local charges, dispersion energy, and the target – receptor distance is preserved. In addition, affinity values for charged targets can be determined. This method evaluates only non-covalent interactions; however, the method can be expanded by including subunits capable of forming specific covalent bonds with selected functional groups of the target in the virtual receptor. Module 5 reflects an example of a preferred affinity calculation scheme used in the present invention.

(5) Assessment of selective affinity

A high degree of correspondence between the virtual receptor and a number of target targets is estimated by comparing the values of the known activity or affinity for the targets with the values obtained for the virtual receptor-target complex. The maximum affinity values of the optimal virtual receptor should be correlated with known affinity values. Successive repetitions of the evolutionary process can be used to increase the degree of this correlation (Fig. 3).

Known values can be any indicator that depends (or is assumed to depend) on the affinity for binding, including (without limitation) EI ₅₀ , GO ₅₀ , affinity for binding, and cohesion values. The test values should be positive. Logarithmic data transformation may be required. You cannot use unreasonably ranked data.

The optimal target orientation to achieve maximum binding affinity is unknown prior to the experiments. In order to obtain a representative value of the receptor affinity for the target, each target should be retested using various random orientations relative to the surface of the receptor. In each experiment, Module 4 is used to evaluate affinity. In general, the reliability of the obtained maximum affinity values depends on the size of the sample, since it increases the likelihood that the sample will have a true maximum value. The same set of target orientations is used to test each receptor.

This method uses two techniques in order to avoid the need for a large set of samples to generate optimized receptors: 1) using a measurement method that combines the average (or total) affinity and maximum affinity for selecting receptors with higher selectivity; and 2) an increase in increments in the number of orientations studied with successive changes in the optimization process (optimization begins with a small number of target orientations, as soon as receptors with a high degree of correspondence are generated, a greater number of orientations are tested).

In accordance with this method, the sum of the affinity values obtained for all tested orientations of each target is determined. This degree of total affinity is a measure of the average magnitude of affinity between the receptor and the target. At the same time, the maximum affinity is also determined.

A correlation between the known values and how total srodst21 CMV and maximal affinity to receive _FOR g ² and g ^2, respectively _MA. The starting point (0,0) is included in the correlation based on the assumption that target compounds that do not exhibit any activity should have little affinity for the virtual receptor or not have affinity at all. This assumption cannot always be reliable, and in some experiments other quantities may be required.

Correlation of total affinity is a measure of average degree of compliance. If this correlation is significant, and the correlation between the maximum affinity and the known affinity is insignificant, the result suggests that the virtual receptor is not selective, i.e. multiple target orientations can efficiently interact with the receptor. Conversely, if the maximum affinity is largely correlated with known affinity values and the correlation of the total affinity is small, the virtual receptor can be highly selective. If both the total affinity and the maximum affinity are significantly correlated with the known affinity, it is likely that the orientation of the samples identified the response characteristics of the receptor with limited errors (errors of type I and type II are reduced: the similarity of an incorrect or negative result). In some cases, it may be appropriate to minimize the correlation between the known affinity values and the total affinity value, at the same time choosing an increased degree of correlation between the maximum affinity and the known affinity value. Such a choice will require subtracting the maximum affinity from the total to exclude these values as ambiguous.

According to this method, the value of the combined correlation based on the selection selection is used. This value is defined as the square root of the product of total affinity and maximum affinity

P = (Tma ² X T ₃ A ² ) ⁰ · ⁵

This value is optimized in the evolutionary process used for virtual receptors. Note: If r _MA ² and rza ^{2 are} correlated to a large extent with each other, then the values affecting rza ² should either individually correlate with the maximum affinity or slightly affect the total value. In addition, a correlation (g z _-ma ) can be calculated for the dependence of the value (total affinity - maximum affinity) on the known affinity value and the criterion

P = (Gma ² X (1-Tza-ma ² )) ⁰ ' ⁵ maximized. Using this criterion allows you to select receptors that have a high affinity for a very limited set of target orientations. Module 5 illustrates a scheme for calculating the degree of conformity of a sample.

(6) Optimization process

The goal of the optimization process is to identify a virtual receptor that has a selective affinity for a set of targets. A highly efficient mechanism for finding solutions is required, since the total number of possible genotypes containing 300 teams is 7 or about 10. The following four phases summarize the stages of the optimization process, with each phase being discussed in more detail with examples of calculations.

Phase 1: Generate a set of arbitrary genotypes and search for a minimum level of activity. Use the selected genotype as a basis for further optimization using a genetic algorithm (recombination) and a unidirectional mutation technique.

Phase 2: Carry out a mutation of the selected genotype to obtain a breeding population that differ from each other, but related genotypes for recombination. Select the most selective mutants from the population for recombination.

Phase 3: Generate new genotypes by recombination of selective mutants. Choose from the obtained genotypes those that show the highest degree of compliance. Use this subpopulation for subsequent recombinant or mutational generation.

Phase 4: Select the best recombination products and repeat point mutations to increase selectivity.

Phase 1: Evolution-generation of primary code.

To find optimal solutions to many problems, the genetic algorithm developed by Ho11apb (Ho11apb, TN. (1975) Abarlaiop ίη Иa1iga1 apb ArPPs1a1 8u§1ssh5. And. Myuyapd Rte88. App Atyoit) can be used. Typically, this technique is used for a large number of initially arbitrary decisions. When implementing this method, the technique was significantly changed in order to reduce the number of tests and iterations required to find virtual receptors with high selectivity. This change was made by using a set of related genotypes as the initial population and applying high mutation rates at each iteration. For any set of target compounds, various receptors with optimal affinity values can be created. For example, receptors can optimally bind to the same targets, but in different orientations. Using the initial population of closely related genotypes increases the likelihood that the optimization process will be reduced to one solution. The recombination of unrelated genotypes, although it may lead to the formation of new genotypes with an increased degree of correspondence, is more likely to lead to different solutions.

The goal of the first stage of the optimization process is to obtain a genotype with a minimum degree of affinity for a set of targets. This genotype is then used to generate a population of related genotypes. The scheme of the genotype generation process with a minimal degree of affinity is reflected in Module 6.

Phase 2: Evolution-mutation of the primary code.

A mutation of a genotype consists in changing one or more features of a code. According to this method, mutations do not lead to a change in the number of subunits forming the receptor polymers and do not affect the length of the genotype. It is clear that these conditions are arbitrary and it is clear that in some systems other options may be used.

Mutations can change the pattern of phenotype folding, leading to a change in the spatial shape of the receptor and the location or exposure of the binding sites. Mutations that affect the configuration of the peripheral regions of the phenotype can lead to a shift in the center of the receptor relative to the center of the target.

Neutral mutations.

All types of mutations lead to a change in the structure of the phenotype, however, not all cause a change in the functionality of the receptor. Such neutral mutations can cause changes in the components of the receptor that do not affect the magnitude of the affinity. In some cases, these neutral mutations may combine with subsequent mutations, resulting in a synergistic effect.

Population for reproduction.

The purpose of the second phase of the evolutionary process is to generate different but related genotypes derived from the primary genotype. The members of this population are then used to generate recombinants. This breeding population is created by multiple mutations of the primary genotype. The obtained genotype is translated and screened to achieve selectivity. Products with the highest selectivity are stored for recombination. Module 7 reflects a diagram of the process of multiple genotype mutations.

Phase 3: Evolution-recombination.

The purpose of recombination is to create new genotypes with an increased degree of correspondence. Recombination facilitates the preservation of fragments of the genotype, which are essential for phenotype matching, and at the same time it leads to the introduction of new combinations of teams. In general, recombination in combination with selection results leads to rapid optimization of selectivity. Module 8 illustrates a diagram of the process of recombination of a genotype.

This method provides for the preservation of the population used for recombination for testing in stage 7 of Module 8. This leads to the fact that genotypes with high selectivity are not replaced by genotypes with lower selectivity. In addition, mutations (Module 7) are used for 50% of the recombinant genotypes before testing (stage 7 Module 8). This stage leads to variability of the recombinant population. Test populations according to this method can have from 10 to 40 genotypes. This population is relatively small. In some circumstances, larger populations may be required.

Phase 4: Evolution-maturation.

The method of progressive micromutation.

The final stage of the optimization process mimics the maturation of antibodies in the mammalian immune system. A series of single point mutations are applied to the genotype and the effect on phenotype compliance is evaluated. In contrast to recombination, this process usually leads only to small incremental changes in the phenotype selectivity. The ripening process is used in the evolutionary strategy of (1 + 1) Rechenberg (ResNepbegd, I. (1973), Euphorbiaceae r. Rottottai. ΔΙηΙΙ give). With each generation, the correspondence of the precursor genotype is compared with the correspondence of its mutation product, and for the next generation, the genotype is preserved with greater selectivity. As a result, this process is strictly unidirectional, since less selective mutants do not replace their predecessors. Module 9 reflects the outline of an example of genotype maturation (this example is not limiting).

During each iteration of the ripening process, only one code command changes. If the precursor and the product of its mutation have the same selectivity, the precursor is replaced by its product in the next generation. This method leads to the accumulation of neutral mutations, which can give a synergistic effect in combination with subsequent mutations. This condition is arbitrary.

If recombination or maturation does not result in improved selectivity after repeated iterations, it may be necessary to repeat phase 2 in order to increase the variability of the genome of the selection population.

Some applications of this method

The method of this invention can be used in several areas, including 1) screening for compounds with selected pharmacological or toxicological activities25; and 2) the creation of new chemical structures with selected functional characteristics. The following are examples of this application.

1A) Screening Method

A population of receptors that have been identified as having selective affinity for a specific group of compounds having the same pharmacological properties can be used as probes to identify other compounds with the same activity, provided that this activity depends on the binding affinity. For example, a population of receptors with a specific affinity for salicylates can be detected. If the affinity of these receptors for salicylates is closely correlated with the affinity of cyclooxygenase for salicylates, the receptors should at least partially mimic functionally relevant features of the binding site of the cyclooxygenase molecule. These receptors can therefore be used to screen other compounds with possible affinity for binding to cyclooxygenase.

This technique can also be used to screen compounds that are potentially toxic or carcinogenic. For example, receptors that mimic the affinity for specific binding of steroid hormone receptors can be detected. These receptors can then be used to assess the affinity of pesticides, solvents, food additives, and other synthetic materials for possible binding affinity prior to the ίη νίίΓΟ or ίη νίνο tests. Simulated receptors can be created that show affinity for alternating target sites transported by protein or not binding to targets.

1B) Screening for submaximal activity

In some cases, compounds with high affinity may exhibit harmful side effects or may not be suitable for administration to chronic patients. In this case, compounds with a lower binding affinity may be required. Techniques such as combinatorial synthesis do not provide for the easy generation or identification of such compounds. In contrast, simulated receptors can be used to effectively screen structures that exhibit binding affinity at any level.

1C) Measurement of molecular similarity

The selectivity of simulated receptors can be used as a quantitative measure of molecular similarity.

An example of simulated receptors.

In this example, to demonstrate the ability of a receptor generation program to produce simulated receptors that mimic any random activity, fictitious target affinity tests were selected.

In this example, all receptors are composed of 15 polymers. The values of width, length and depth indicate the initial coordinates of 1 5 polymers relative to the center of the receptor.

Example 1

A simulated receptor was obtained with the following characteristics:

number of subunits: 240; width: 6; length: 6; depth: 25

The code:

4100033103212204103333424052312013341024 12402223233401003224251014405133243400324 62041210013131004311210113241202242130241 32311243301331003230523000433414010202230 2140414143102103103102103103202103103103103103103304304304304304303303304303104304103103404303103403403403403403403403403303303303303303103103000

Each target was tested 20 times with respect to the receptor.

The affinity of the optimized receptor is 0.9358, i.e. relatively low.

The target substrates used to optimize the receptor were benzene, phenol, benzoic acid, and osalicylic acid. The aspirin precursor, o-salicylic acid, is an inhibitor of prostaglandin synthesis by cyclooxygenase. Benzoic acid and phenol have a much lower affinity for the same site. The affinity of the target and the degree of affinity for the receptor are shown below in table. A, which shows that the simulated receptor has a maximum affinity for o-salicylic acid.

Table a

Compound- target Affinity target Power total affinities Power maximum affinities Benzene 0.6 20.88 3.38 Phenol 1,2 8.03 4.99 Benzoic acid 1,6 41.23 12.98 o-salicylic acid 4.4 80.33 34.71

Three test substrates were evaluated using a simulated receptor. Two compounds are known to be less active than o-salicylic acid: msalicylic acid and p-salicylic acid. The third compound, ISh81pa1, is a derivative of fluorinated salicylic acid, the effectiveness of which is equal to or higher than the effectiveness of salicylic acid. The evaluation results are given in table. IN.

Table B

Target compound Power total affinities Power maximum affinities m-salicylic acid 45.9 12.3 p-salicylic acid 63.5 27.5 IIA8sha1 117 71.2 o-salicylic acid 80.33 34.71

The results obtained using the simulated receptor correspond to the pharmacological data for these compounds: m-salicylic acid and p-salicylic acid are characterized by lower affinity levels than o-salicylic acid, and ύίΠιϊδίηαΙ is more active than o-salicylic acid. Further purification of the simulated receptor and the use of additional, independently optimized receptors will be required to increase the accuracy of these predictions of activity.

Part B. Creating new compounds with selected functional properties

The evolution of new ligands.

A population of simulated receptors with selective affinity for a set of target compounds with the same functional properties can be used to create new compounds with the same properties, provided that these properties are closely correlated with the structure or binding affinity of model compounds. Using interaction with receptors as a selection criterion, it is possible to identify new chemical structures that are optimal for receptors. Since these compounds must meet the necessary and sufficient requirements for receptor selectivity, it is likely that these new compounds also have activity similar to the activity of the original molecular targets.

The general scheme of the process.

1. Obtain a population of simulated receptors with optimized selectivity for the set of characterized target compounds. In some cases, it may be desirable to generate several populations with different affinities. For example, three populations of simulated receptors can be obtained, the first simulating the properties of a selected target site, the second simulating the site required for transporting a ligand to its main target, and the third a population of simulated receptors simulating a target site causing undesirable side effects. The creation of a new ligand structure in this case will require simultaneous optimization of affinity for the first two receptor populations and minimization of affinity for the third population.

2. Determine the affinity of the new basic structure for the population (s) of simulated receptors.

3. Modify the basic structure and evaluate the affinity using the population (s) of simulated receptors. If the modification improves affinity, the modified structure is retained for further modification. Otherwise, another modification is tested. Previously rejected modifications can be reused in combination with other modifications.

4. Step 3 is repeated until a compound with suitable affinity values is obtained.

Note: Using discriminatory simulated receptors, it is possible to identify chemical structures with an affinity for the selected target site below the maximum.

1) Molecular genotype code generation

Encoding the ligand phenotype (molecular structure) as a linear genotype, represented by a string of characters (characters), facilitates the processes of mutation, recombination, and inheritance of the structural characteristics of the ligand during the evolutionary process.

Ligands detected by this method are composed of substituted carbon skeletons. Each code consists of three feature vectors. The primary code vector contains varying instructions for generating the carbon skeleton and determines the position of each carbon atom in the skeleton. The secondary code vector identifies functional groups attached to each carbon atom. The tertiary code vector determines the position of the functional group with respect to the carbon host. Molecular skeletons connecting atoms other than carbon (for example, in ethers, amides and heterocycles) can be created homologously using additional symbols in the code to designate atoms replacing carbon atoms in the skeleton.

The carbon skeleton is created from a series of points that form the nodes of a three-dimensional tetrahedral coordinate system. During the creation of the carbon skeleton, the distance between the nearest points is equal to the average bond length between the carbon atoms of the alkyl.

Primary code vector: determinants of the ligand skeleton.

The primary code vector consists of characters identifying a changing command with respect to a given atom position. Each changing command determines the coordinates of the next atom in the tetrahedral matrix. Four directions can be taken from each atom (1, 2, 3, 4), corresponding to unfilled valences of 8p ³ carbon. Each of the carbon atoms belongs to one of four possible states (A, B, C, I). These states correspond to the number of nodes in the tetrahedral coordinate system.

The relationship between the direction of rotation and the new coordinates for the next atom in the skeleton is shown in the following tables. The two tables B1 and B 2 below represent the two changing conditions required to create the ligands. The condition for the formation of a bath conformation leads to the generation of a tetrahedral matrix in which closed 6-membered rings (cyclohexanes) accept the bath conformation. The condition for the formation of chair conformation leads to the generation of a matrix in which cyclohexyl rings accept the chair conformation. You can combine both conditions during code generation. In the examples discussed in this description, only the condition for the formation of conformation of the chair is used.

Table B1

Conditions for the formation of a bath conformation Current position = (x, y, ζ) New position after rotation

The present position Turn = 1 Turn = 2 Turn = 3 Turn = 4 BUT (x-.75, y + .433, ζ-.5) (x-.75, y + .433, ζ-.5) (x, y-.864, ζ-.5) x, y, ζ + 1) IN (x + .75, y-.433, ζ + .5) (x-.75, y-.433, ζ + .5) (x, y + .864, ζ + .5) x, y, ζ-1) FROM (x-.75, y + .433, ζ + .5) (x + .75, y + .433, ζ + .5) (x, y-.864, ζ + .5) x, y, ζ-1) AND (x + .75, y-.433, ζ-.5) (x-.75, y-.433, ζ-.5) (x, y + .864, ζ-.5) x, y, ζ + 1)

Each rotation leads only to determine the state of the new atom:

New state after turning

The present position Turn = 1 Turn = 2 Turn = 3 Turn = 4 BUT IN IN IN FROM IN BUT BUT BUT AND FROM AND AND AND BUT AND FROM FROM FROM IN

Table B2

Conditions for the formation of chair conformity Current position = (x, y, ζ) New position after rotation

The present position Turn = 1 Turn = 2 Turn = 3 Turn = 4 BUT (x-. 75, y + .433, ζ-.5) (x-.75, y + .433, ζ-.5) (x, y-.864, ζ-.5) x, y, ζ + 1) IN (x + .75, y-, 433, ζ + .5) (x-.75, y-.433, ζ + .5) (x, y + .864, ζ + .5) x, y, ζ-1) FROM (x-.75, y-, 433, ζ + .5) (x + .75, y-.433, ζ + .5) (x, y + .864, ζ + .5) x, y, ζ-1) AND (x + .75, y + .433, ζ-.5) (x-.75, y + .433, ζ-.5) (x, y-.864, ζ-.5) x, y, ζ + 1)

Each rotation also leads to the determination of the position of the new atom:

New state after turning

The present position Turn = 1 Turn = 2 Turn = 3 Turn = 4 BUT IN IN IN FROM IN BUT BUT BUT AND FROM AND AND AND BUT AND FROM FROM FROM IN

Using these relations, it is possible to decode the primary code vectors, consisting of strands of symbols 1, 2, 3 and 4 with the creation of three-dimensional formations of carbon atoms. The resulting filaments of carbon atoms allow folding or forming closed loops to produce short side chains and ring structures. Specific ring structures (e.g. cyclohexanes) can be introduced directly in the form of representative character sequences, as shown below.

Secondary code vector: substituents. A secondary code vector, of the same length as the primary code vector, is used to highlight the type of substituent at the carbon atom defined by the primary code vector. Each substituent is identified by a separate symbol. Substituents are added one to the carbon skeleton. A single carbon atom may have more than one substituent, but only if it is described by the primary code more than once.

In the process of constructing the ligand in this method, all valencies not filled with substituents described by the secondary code vector are automatically saturated with hydrogen atoms. Other rules can be used to saturate free valencies with atoms other than hydrogen.

Tertiary vector: vector of communication with a substituent.

A tertiary code vector, of the same length as the primary code vector, is used to highlight the valency linking the substituent described by the secondary code vector. The tertiary code consists of characters 1, 2, 3 and 4, each of which refers to the direction of rotation defined for the primary code. Substituents are distributed only if the valency is not yet filled either with a carbon atom defined by the primary code vector, or another, previously allocated (distributed) substituent. Or, subsequent substituents may replace previously distributed substituents.

2) Creating a program (code)

To create carbon skeletons, a primary program (primary code) is constructed using random ordering (sequence) of characters belonging to the set (set) {1, 2, 3, 4}. The creation of heterocyclic structures, ethers, amides of imides and carboxylic compounds is carried out by replacing the skeletal carbon atom with another atom defined by the secondary code (secondary program).

The secondary code is created based on the random ordering of characters that determine the types of substituents. The frequency of characters may be random or previously fixed before generating commands.

The tertiary code consists of characters from the set (set) {1, 2, 3, 4}. Cyclic structures can be constructed consciously (as opposed to random generation) by adding (sequences) the ordering of specific characters to the primary code. For example, 431423 encodes cyclohexyl rings. A sum of 24 lines encodes all possible positions of cyclohexyl groups in the tetrahedral matrix. The secondary and tertiary code vectors for the primary loop codes are created as described above. Module 10 is a block diagram of an exemplary program for generating carbon skeletons including cycles.

The relative positions of the entry and exit points of the part of the carbon skeleton containing the cycle is determined by the length of the orderings of the symbols used to create the cycle. Specifically, if the ordering (sequence) contains six characters, for example, 432413, then the entry and exit point is the same member of the cycle. If the sequence is partially repeated and added to the original six characters, the entry point and exit point will not be the same member of the loop. For example, 4314134 and 43141343141 form cycles in which the exit points of the ring members are adjacent to the entry points.

In this case, the rings are attached to the skeleton, adding sequences of 6 or more characters to the code. For the cycle described by 431413, the possible sequences used (ordering) are:

431413

4314134

43141343

431413431

4314134314

43141343141

431413431413

431413431413

The proposed rules for creating a new ligand genotype can be used to encode other chemical structures in a linear format both for storage (in memory) and for introducing ligands into the evolution process. For example, a well-known pharmacophore can be encoded in a linear format and used as a starting point to identify new ligands with similar or improved functional properties. Similarly, pharmacophore sets interacting with a common target site can be encoded in a linear format and used for recombination.

3) Code translation and ligand construction (construction)

Code vectors are converted into three-dimensional images of ligands in the translation process, which consists of three discrete stages. At the first stage (construct), a carbon skeleton is built using the primary (main) code. In the second stage, the carbon skeleton is supplemented with substituents using the commands of the secondary and tertiary code vectors. The teams of the vectors of the secondary and tertiary code can also determine (establish) the substitution of carbon atoms in the skeleton for other atoms. The teams of the secondary and tertiary codes can also change the number and direction of valencies of a carbon atom or another atom that is part of the main (primary) skeleton. For example, the addition of carbonyl oxygen saturates (fills) two free valencies. In the third stage, all valencies not filled with substituents in the second stage are saturated with hydrogen atoms (unless otherwise indicated).

Primary decoding: building a ligand skeleton.

Primary decoding uses rotation instructions from the primary code vector to establish the position of each carbon atom. It is assumed that the first atom is located at the origin of the coordinate system. It is assumed that the first atom occupies state A.

Decoding occurs sequentially. The result of the primary decoding process is a 3 x η matrix containing the x, y, and ζ coordinates of each of the η skeletal carbon atoms. Since loops and turns are allowed, more than one carbon atom can occupy the same position in space. In such cases, it is assumed that only one carbon atom occupies a given position. As a result, the number of carbon atoms forming the final skeleton may be less than the number of characters in the primary code vector.

When the primary code is read, a list (list) is created from the secondary code that defines the substituents for each carbon atom. At the same time, a parallel list is created using a tertiary code to determine the valency saturated by each substituent.

Secondary decoding: the introduction (addition) of substituents.

Substituents are added (introduced) sequentially to each carbon atom based on a list formed from the secondary code during primary decoding. To determine the position of the valency of the substituent relative to the basic (skeletal) carbon (carbon host), use the corresponding value from the tertiary code. If the position is already occupied either by an adjacent carbon atom or by a substituent previously described, substitution is not performed. Or, a decoding process can be created in which the substitution is made to the next unoccupied position, or the previously described substitute is replaced. The distance between the substituent and the carbon atom is calculated from reference tables of bond lengths. These positions (positions) and bond lengths are used to calculate the substituent coordinates. In the case of multicomponent substituents, for example, hydroxyl, nitro and amino groups, coordinates for the carbon host are calculated for each atom in the substituent.

After all the substituents described using the secondary code vector are introduced into the skeleton, all the unfilled positions remaining in the skeleton are saturated with hydrogen atoms. To calculate the coordinates of each hydrogen atom, the hydrogen bond length is used. 3 §p is a carbon atom.

Individual carbon atoms may have more than one substituent other than hydrogen. This can happen if the same position is described by the primary code vector more than once. This embodiment of the invention does not include multiple substituents directly using a secondary code, although this can easily be done.

Substituents are allowed only in positions not occupied by carbon atoms forming the ligand skeleton. The summary (cumulative) list contains all occupied sites (positions) in the tetrahedral matrix.

In the process of secondary decoding, a list of types, radii, and positions of all atoms contained in the ligand is translated. This list is the basis for the subsequent generation of targets.

At this stage of the process, the feasibility of a structure created on the basis of ordering (sequence) of code is not evaluated. In some cases, atom coordinates can be entered into energy minimization programs to create more plausible structures. However, in this case, no assumptions are made regarding the configuration of the ligand during binding. In addition, this embodiment of the method preserves the uniqueness of the structure of a specific configuration of the same molecule. For example, this development distinguishes between three rotational isomers of butane and considers each isomer as a separate molecule.

The vectors of the codes make up the genotype of the corresponding ligand and can undergo (change) mutation and recombination, resulting in a change in the structure of the ligand. The ligand structure itself is a phenotype used to evaluate binding affinity for a selected population of virtual receptors.

4) Presentation of targets

The chemical structures of ligand targets are initially built on the basis of randomly generated (created) codes. After decoding, from the look-up tables, the coordinates, radii, dipole moments, and polarizabilities of each atom in the target ligand are obtained and used to assess the binding affinity between the ligand and the selected virtual receptor population.

The affinity of the target for each of the virtual receptors is checked for many orientations of the target relative to the surfaces of the receptor. No assumptions are made regarding the relative orientation of the ligand and simulated receptor. Before evaluating binding affinity, the target and receptor must come into contact. The method for presenting a target and calculating the affinity between chemical structures and simulated receptors is almost the same as that discussed above in Module 4 between molecules of known targets and simulated receptors.

5) Assessment of binding affinity and compliance

The affinity for binding the target ligand to each of the simulated receptors used to (evaluate) the determination of conformity is calculated using the same method for calculating the effective affinity that is described for generating simulated receptors using target molecules. As noted earlier, affinity calculations can be introduced into the conformity checking process using other criteria, but the efficiency and accuracy of the calculations in this case depends in part on the use of the same effective affinity calculation for generating virtual receptors and generating chemical structures using populations of simulated receptors.

6) Change (evolution) of the ligand

Check compliance.

The degree of correspondence between the selected population of simulated receptors and a new ligand or chemical structure is evaluated by comparing the activity or affinity of the target to the ligand with those obtained for simulated receptor-ligand complexes. The maximum affinity values of the optimal selective virtual receptor should clearly correlate with the affinity values of the target. Subsequent operations of the evolution (change) process are used to increase the degree of this correlation (correspondence).

The affinity for binding to the target can (be at any level) be different. It is not required that the ligand have the same affinity for binding to all virtual receptors during the selection process. In this case, the maximum affinity for binding of optimized virtual receptors to known substrates is used to calculate the target affinity for binding. For example, target affinity values can be up to 90% of the affinity for binding each member of the virtual receptor population to a particular substrate. Or, the binding affinity of the target may be near zero if the interaction between the ligand and the virtual receptor should be minimized.

By combining simulated receptors optimized for different sets of substrates and associating selected target affinity values with each receptor, new ligands with specific binding affinity parameters can be selected. Ligand compliance is a measure of how the values of the binding affinity calculated for the ligand and the affinity of the target match. The optimization process maximizes ligand compliance.

The optimal orientation of the ligands in order to achieve maximum affinity for binding prior to testing is unknown. In order to obtain a representative order (size) of the receptor-ligand affinity interval, each new ligand needs to be checked repeatedly for various random orientations relative to the plane (surface) of the receptor. Each test uses Module 4, discussed in Part A, to evaluate affinity. In general, the reliability of the obtained maximum affinity values depends on the size of the sample, since the probability that the sample contains a truly maximum value increases.

To avoid the need for a large sample to generate optimized new ligands or chemical structures, two techniques are used in this aspect of the invention:

1. The use of a value that combines the average (or total) affinity and maximum affinity, in order to select ligands with optimized affinity profiles.

2. Incremental increase in the number of orientations verified by successive iterations of the optimization process. (Optimization begins with a small set of target orientations; when more suitable ligands are generated, more orientations are tested).

In this case, the sum of the affinity values obtained for all verified orientations of each ligand is calculated. This total affinity value is a measure of the average affinity between the receptor and the ligand. At the same time, the maximum affinity is also determined.

Both total and maximum affinity are used to check the degree of correspondence of the virtual receptor and the new ligand. The correspondence of each new ligand is evaluated based on the difference between the calculated values of the total and maximum affinity and the values of these parameters for the target. In this case, the value:

calculated as the degree of compliance for each new ligand-modeled receptor pair. Compliance is maximum when the degree of compliance is zero. The maximum affinity and total affinity of the targets were obtained using inverse functions found during the evolution of optimized virtual receptors, as described in the previous sections. The values for the targets are obtained as follows:

Max. affinity of the target = £ x maximum affinity of the most active (strong) substrate used to generate the virtual receptor; total affinity of the target = £ x and total affinity of the strongest substrate used to generate the virtual receptor, where £ = scaling factor.

When more than one simulated receptor is used to evaluate ligand conformity, the degree of compliance of each ligand-simulated receptor pair is summed.

In this case, the degree of compliance is maximized when the sum of the degrees of compliance is zero. In some cases, it is desirable to use only the degree of maximum affinity when testing a new ligand relative to a panel of various simulated receptors. In this case, compliance will be expressed as:

P (D (= £ 'max calculated affinity, - max, the affinity for the target ^{1/1 = 1} max, affinity for the target _{

In this case, the correspondence is also maximized when the sum of the degrees of correspondence is zero. Other methods, for example, a geometric method, can be used to measure the overall degree of compliance of the ligand being tested against a number of simulated receptors.

Using values of both maximum affinity and total obtained for each simulated receptor ensures that the selectivity of virtual receptors is included in the assessment of ligand compliance. Thus, the ligand match reflects not only the ligand affinity, but also the fact that it satisfies the spatial requirements of the virtual receptor, which underlie selectivity.

6a) The purpose of the optimization process

Create a new ligand with selected affinity for the target for a set of simulated receptors. A highly effective mechanism is required to solve (this problem), since the total number of possible genotypes corresponding to 25 teams is 256 ²⁵ .

Process.

(1) Phase 1. Create a set of random genotypes encoding the ligands, and screen for a set of simulated receptors to select ligands that exceed the degree of compliance threshold.

(2) Phase 2. The selected genotype is used as the basis for further optimization, using the methodology of the genetic algorithm (recombination) and unidirectional mutation. Mutate the selected genotype to generate a population to reproduce individual but related genotypes for recombination.

(3) Select the most selective mutants from the population for recombination.

(4) Phase 3. Create new genotypes by recombination of selective mutants. Select from the obtained genotypes those corresponding to the highest degree of affinity. Use this subpopulation for subsequent generation of recombinations (repeat phase 3) or mutations (repeat phase 4).

(5) Phase 4. Take the best recombination products and apply repeated point mutations to them to increase selectivity.

(6) The optimization process is completed when ligands with a given match are generated (created).

Phase 1: Evolution-Generation of Primary Code

The task of the first stage of the optimization process is to create the genotype and phenotype of the corresponding ligand with a minimum degree of compliance. This genotype is then used to create a population of related genotypes.

The genetic algorithm discovered by Ho11aib (Holland) can be used to search for optimal solutions to many problems. Typically, this technique is applied using large, initially random sets of solutions. In this aspect of the invention, the technique is significantly modified to reduce the number of tests and iterations required to find ligands with high selective affinity. This was accomplished using a set of genotypes with close affinity as the initial population and the use of high mutation rates at each iteration. For each set of target compounds, one can obtain (find) individual ligands with optimal affinity characteristics. For example, receptors can optimally bind to the same targets, but with different orientations. The use of the initial population of closely related genotypes increases the likelihood that the optimization process will be reduced to a single solution. Recombination of unrelated genotypes, although it may create new genotypes with increased compatibility, it is more likely that it will lead to divergence.

Phase 2: Ligand Mutation

The task of the second phase of the evolution process is to generate a population of separate but related genotypes formed from the primary genotype. Members of this population are then used to obtain recombination. This population for reproduction (improvement) is created using the multiple mutations of the primary genotype. The resulting genotypes are translated and screened for selectivity. Products with the highest selectivity are stored for recombination.

Ligands undergo mutations by changing the characteristics (symbols) of the genotype (code vectors) encoding their structures. These mutations change the size of the ligand, as well as the location and type of the functional group present in the ligand. Mutations in this case can change the number of carbon atoms that make up the ligand skeleton. Module 11 is a flowchart of an exemplary multiple point mutation process.

Mutations can change the nature of the ordering of the ligand phenotype, resulting in a change in the size and location or appearance of the functional group. Mutations that affect the configuration of the peripheral regions of the ligand phenotype can lead to a shift in position relative to the receptor center.

Neutral mutations.

All mutations change the structure of the phenotype, however, not all mutations lead to a change in the functionality of the ligand. Such neutral mutations can change the components (components) of the ligand, which do not affect the affinity. In some cases, such neutral mutations can be combined with subsequent mutations, providing a synergistic effect.

Mutation Sequences.

Sequence mutations do not directly change code characters. Instead, the sequence of characters in the code is changed. Mutation sequences can change the size of the ligand, (structural) configuration and the presence and location of functional groups. Four types of sequence mutations are used in this aspect of the invention:

a) DELETION: a sequence of characters is removed from the code.

AVSEEA AVEA

b) INVERSION: the order of the characters that make up the sequence within the code is reversed.

AVSEEA AVESEA

c) DUPLICATION (doubling): the sequence of characters that make up part of the code is repeated.

AVSEEEAAVSESEEEA

L) INSERT: a sequence of characters is embedded in the code.

AVSEEA AVSEEVSEA

The mutations of this invention are used in combination. Module 12 is a block diagram of an exemplary sequence mutation.

Phase 3: Recombinant Code Generation

During recombination, randomly selected, complementary sections are exchanged between the selected genotypes. The task of recombination is the generation of new genotypes with increased compliance. Recombination facilitates the conservation of genotype fragments that are essential for phenotype matching, and at the same time introduces new combinations of commands. In general, recombination combined with selection gives a quick optimization of selectivity. Module 13 presents a block diagram of an exemplary recombination procedure.

In this case, the population used for recombination for the purpose of testing is preserved. This ensures that genotypes with high selectivity are not replaced by genotypes with a lower match. In this aspect of the invention, up to 50% of the recombinant genotypes undergo multiple mutations before testing. This process increases the variability within the recombinant population. The tested populations used according to this invention have sizes ranging from 10 to 40 genotypes. This is a relatively small population size. In some cases (under certain conditions) large populations may be required.

Phase 4: Ligand Maturation

The method of progressive micromutation.

At the final stage of the optimization process, antibody maturation in the mammalian immune system is simulated. A series of single-point mutations are carried out in the genotype and the effect on phenotype compliance is evaluated. Unlike recombination, this process usually leads only to small changes that increase the selectivity of the phenotype. The ripening process uses the (1 + 1) Rechenberg evolutionary method (KesNeuber). With each generation (generation), the correspondence of the initial genotype is compared with that of the product of the mutation, and the genotype with higher selectivity is retained for subsequent generation. As a result, this process is strictly unidirectional, since less selective mutants do not replace the precursor. At each iteration of the ripening process according to this invention, only the only command in the code changes.

If the precursor and the product of its mutation have the same selectivity, the precursor is replaced by the product in the next generation. This method leads to the accumulation of neutral mutations, which may have a synergistic effect with subsequent mutations. This condition is optional, random. Module 1 to 4 is a flow chart of an exemplary ripening process.

If recombination or maturation does not provide improved selectivity after repeated iterations, it may be necessary to repeat multiple mutations (phase 2) in order to increase the variability of the genome of the population for reproduction (reproduction).

Ligand Generation Examples

General view

Mosquito Lebeck Idur (yellow fever) repels benzaldehyde and, to a much lesser extent, benzene and toluene (Table 1). This species does not significantly deter cyclohexane or hexane (Table 1). In the next test of the new generation of ligands, a method is used to create, a ίηίίίο (from the beginning), compounds that would have activity similar to benzaldehyde as repellents. At the first stage of ligand generation, simulated receptors with high affinity for benzaldehyde and low affinity for benzene were created. At the second stage, ligands with an affinity for binding to simulated receptors similar to that of benzaldehyde are detected.

Reactions (response) of mosquitoes

Mosquitoes are laboratory grown, 7-14 days old and not fed. The experiments are carried out for six days at 20 ° C and ultraviolet radiation. Tests are carried out between 12.00 and 17.00 Eastern Standard Time. Test populations in four test kits contain 200, 175, 105, and 95 females. Mosquitoes are given drinking water.

Tests are carried out in cases (boxes) of transparent plexiglass measuring 35 x 35 x 35 cm with two opposite sides, which are screen screens. Screening (screening) passes through two layers: the inner plastic layer with large holes and the outer nylon layer with frequent (small) holes. The pencil case is placed in a vapor atmosphere so that air enters through one of the shielded sides and exits through the opposite side. Air current <0.5 cm / s.

Mosquitoes are placed on the walls of the pencil case (box) with their heads up. To supply a stimulating substance, triangular (4 x 4 x 1 mm) paper filters are used (XV On I and η, Whatman, No. 1). The ends are immersed in the test solution to a depth of 0.5 cm and used immediately. The reaction (response) to the test solutions is determined as follows:

. A fixed female, located on the inner sieve of the windward wall, is selected for testing.

2. The processed end of the paper filter is placed on the outside of the screen sieve and placed opposite the mid-thoracic paw of the mosquito. In all cases, start the approach from the bottom of the mosquito.

3. The tip (filter) is kept in (this) position for a maximum of 3 s and the mosquito reaction is noted.

Then repeat the procedure for a new individual. Mosquitoes are tested only once daily for each compound. The tips (filters) are used for five tests each (total duration of use <30 s), then replaced. The connections are checked at random, and each connection is checked twice on different days. Two sets of control (experiments) are carried out using the untreated (dry) ends of the filter paper and the ends moistened with distilled water, in the same way as with the treated ends. Tests of these control samples are regularly distributed among repellent tests. The reaction to control compounds does not change during the experiment (p <0.25).

Four behavioral (behavioral) reactions are noted (recorded):

. No reaction: the mosquito remains motionless.

2. Take-off: a mosquito flies away from its location.

3. Raising the ipsilateral leg (tarsus): the mosquito raises the mid-thoracic tarsus on the same side as the source of the stimulus.

4. Raising the contralateral paw: the mosquito raises the mid-thoracic paw on the side opposite to where the source of the stimulus is located.

The rise of the ipsilateral paw is often followed by take-off, in which case both types of behavior are recorded. During testing and at all phases of the preparation of the compound, polyethylene gloves are used.

Table E1

Mosquito response to selected volatile compounds

Compound Boiling point, ° С N % Reaction- takeoff % Reaction raising the paws Relative repulsive ability * Benzaldehyde 178 130 90 10 178 Benzene 80 72 72 12.5 68 Toluene 110 166 67 27 94 Cyclohexane 81 80 6 0 4.9 Hexane 69 one hundred 4 0 2.8 Control (blind) 450 5 0

* Relative repulsive ability = [(% Reaction-take-off +% Reaction of the foot) H boiling point / 100

Modulated receptor and ligand generation.

Simulated receptors generate (receive) using the same selection criteria. Each receptor is used independently to produce a set of ligands.

Selection of molecules 1

Phase 1: receptor generation.

A receptor with selective affinity for benzaldehyde is detected. The test targets are benzene and benzaldehyde. Fifteen orientations of each target are used to calculate affinity values.

The results of the evolutionary process are as follows:

Target Value activity Total affinity Maximum affinity Benzene 1,0 6.87 2.21 Benzaldehyde 5.9 75.87 13.02

The degree of affinity for the receptor is 0.992.

Code for the optimized 25 x 6 x 7 receptor bin

231014406145

412061421302

022224-31514

402131114311

053400324221

413231124335

143431012321

010233120331

412100131300

133100333032

341310334122

260214016231

0631121 01132 300043541401 1 01 4141 41 021

Phase 2: ligand generation An optimized simulated receptor is used as a matrix for the evolution of new ligands. Using random mutation and selection, four different ligands are selected. Ligands are selected similar to benzaldehyde.

The ligand affinity values are as follows:

Benzaldehyde Ligand 1.1 S8N17S12ON Ligand 1.2 C8H15C1 Ligand 1.3 C8H13C1 (= O) Ligand 1.4 C13H16ON (= O) Total affinity 75.87 74.03 67.88 72.25 72.94 Maximum affinity 13.02 12.82 15.14 12.58 11.2

Selected ligands from 1. 1 to 1 .4 are shown in FIG. 4b. At least one orientation of each ligand is similar in structure to benzaldehyde.

Selection of molecules 2

Phase 1: receptor generation.

A 25 x 6 x 7 receptor is detected by selective affinity for benzaldehyde. Benzene and benzaldehyde are used as test targets. Fifteen orientations of each target are taken to calculate affinity values.

The results of the evolutionary process are as follows:

Target Power activity Total affinity Maximum affinity Benzene 1,0 25.88 8.53 Benzaldehyde 5.8 162.23 42.74

The degree of affinity for the receptor is 0.996.

The receptor code is as follows:

031264441313

543301114446

224223503403

323030313214

004422243042

210043042311

432003432122

002321144010

223140112054

323431131340

002221221113

000243013133

302122330134 1 300201 201 33 411440003113

Phase 2: ligand generation.

An optimized simulated receptor is used as a matrix for the evolution of new ligands. Four different ligands are selected by random mutation and selection. Ligands are selected similar to benzaldehyde.

The ligand affinity values are as follows:

Benzaldehyde Ligand 2.1 С8Н13С1 (= О) Ligand 2.2 C9H15C1 (= O) Ligand 2.3 С6Н10СЫ (= О) Ligand 2.4 CH „(= O) Total affinity 162.23 182.4 166.5 159.7 156.8 Maximum affinity 42.74 48.97 43.0 39.0 46.5 Power compliance 0.135 0.02 0.05 0.06

Isolated ligands from 2.1 to 2.4 are shown in FIG. 4s At least one orientation of each ligand is similar in structure to benzaldehyde.

Compounds 2.1 and 2.4 are derivatives of substituted cyclohexanone. Ligand 2.2 is 5-chloronone-2,7-dione and ligand 2.3 is 2-cyanohexan-5-one. Ligand 1 .4 contains a fragment corresponding in structure to methylcyclohexyl ketone. Experiment that these ligands are also repellents of cops to check repellent ability in relation to mosquitoes (Table E2).

cyclohexanone, menton, methylcyclohexylketone and octan-2-one (see Fig. 4a) confirm

Table E2

Mosquito response to selected volatile compounds

Compound Boiling point, ° С N % Reaction- takeoff % Reaction raising the paws Relative repulsive ability * Benzaldehyde 178 130 90 10 178 2-Octanon 173 80 82 12.5 162 2-acetylcyclohexanone 225 one hundred 54 24 175 Cyclohexanone 156 134 99 one > 154 Menton 207 110 72 eleven 172 Control (blind) - 450 5 0

* Relative repulsive ability = [(% take-off +% reaction of the foot) x boiling point / 1 00

The method proposed in this description for constructing new chemical structures exhibiting pre-selected functional characteristics or properties has been described only by way of example. For example, the method can be easily put into practice using other known or acceptable values of polarizabilities, dipole moments, covalent radii, and the like. In addition, the block diagrams giving the stages of the process calculation in the modules are for illustration purposes only. For example, the affinity calculation can be carried out using available computing packages using fewer approximations (approximations) than is done in this description. The method for the formation of new chemical structures is based on the first generation of one or more simulated receptors exhibiting a pre-selected affinity for known target compounds with similar functional characteristics and on the use of these receptors to create new structures possessing these characteristics to a given degree. The receptors themselves can be used for other applications, in addition to the formation (generation) of new chemical structures, for example, as a means for screening the pharmaceutical or toxicological properties of known compounds. Thus, it will be understood by those skilled in the art that many changes to the method described herein may not be made outside the scope of the invention.

Table 1

Transitional states and addition coefficients

Old state Joining ratios New state to rotate = Δχ Δy Δζ Right Up Left Down one 0 one 0 2 4 3 5 2 -one 0 0 fifteen 6 one 24 3 one 0 0 one 7 fifteen 22 4 0 0 -one 12 23 14 one 5 0 0 one nine one sixteen 23 6 0 0 -one eleven twenty 10 2 7 0 0 -one thirteen 21 8 3 8 0 -one 0 7 nine 24 14 nine -one 0 0 17 10 5 8 10 0 one 0 6 14 22 nine eleven 0 -one 0 22 sixteen 6 12 12 -one 0 0 eighteen eleven 4 thirteen thirteen 0 one 0 24 12 7 sixteen 14 one 0 0 4 8 eighteen 10 fifteen 0 -one 0 3 17 2 eighteen sixteen one 0 0 5 thirteen 17 eleven 17 0 0 -one sixteen nineteen nine fifteen eighteen 0 0 one 14 fifteen 12 nineteen nineteen 0 one 0 twenty eighteen 21 17 twenty one 0 0 23 24 nineteen 6 21 -one 0 0 nineteen 22 23 7

22 0 0 one 10 3 eleven 21 23 0 -one 0 21 5 twenty 4 24 0 0 one 8 2 thirteen twenty

Formula for the algorithm: Input data (old state, rotation) output data (Ah, Au,

Δζ, new state)

Example: starting position (12, 34, -18); Input data: old state = 10, turn = right:

Output data: new state = 6, Δх = 0, Δу = 1, Δζ = 0; Subsequent position (12, 35, -18).

table 2

Van der Waals Radii

Element N R ABOUT N FROM C1 3 Vg R I Van der Waals Radius (picometers) 110 140 150 150 170 180 180 190 190 200 Relative radius (H = 0.5) 0.5 0.64 0.68 0.68 0.77 0.82 0.82 0.86 0.86 0.91

Adapted from: N.8. Lzasz, 1987. Ryu81sa1 Ogdats Syetugu. Lopdtap Zaepyys apb Tesytsa1, No. \ ν 828 pp.

Table 3

The radius of covalent bonds (PM)

Communication order First N 28 IN FROM N ABOUT R First 88 77 70 66 64 Second 66.5 60 55 Third 60.2 55 Aromatic 3ί R 70 3 C1 First 117 110 104 99 Vg First 114

Adapted from: N.3. Tzzazz (1987).

Table 4

Values of effective dipole moments used for charge assignment

Communication Atom The magnitude of the dipole moment (Debye) CH N +0.35 or + 0.084 * FROM no charge attribution AgS-N N +0.6 FROM -0.366 or no charge attribution = CH N +0.366 FROM -0.6 or no charge attribution C = O ABOUT -2.7 FROM no charge assignment * or +1.35 SOS ABOUT -0.8 SLEEP N +1.5 or +1.7 ABOUT -1.1 SUN ₂ N +1.3 N -1.3 C-YO2 ABOUT -2.0 N +4.0 C n N -3.7 FROM no charge attribution S-S-S 3 in thiophene or dimethyl sulfide + 1.5 * (may be negatively charged in context) C-S = C N in pyridine or CH ₃ —Y = CH ₂ +1.5 or +1.3 Ar-P or C = CP R -1.3 SR R -1.8

Ar-C1 or C = C-C1 FROM -1.7 C-C1 C1 -2.1 Ar-Br or C = C-Br Vg -2.0 S-Vg Vg -2.0 S-1 I -2.0

* Preferable in most cases

Each target atom is completely described by a set of eight values {x _; , y _; , ζ „η, bg ₁ , cr ₁ , b ₁ , α} !, where x ;, y and ζ are the coordinates of the positions with respect to the geometric center of the molecule, r ₁ means the van der Waals radius, bg ₁ means the bond or the covalent radius, c ₁ means the radius of overlap (collision) (= g ₁ +0.5), α! means polarizability and b ₁ means the value of the effective dipole moment.

Table 5

Selected values of the relative effective polarizabilities of the atoms of the selected targets

Atom Context Relative polarizability (α ₁ ) N CH 1,0 N N-11 1,1 N IT 1,1 N 8-N 3.0 * R SR 1.5 * C1 C-C1 4.0 Vg S-Vg 5.8 I S-1 8.9 * FROM C-CH3 3,7 FROM C-CH2-C 3,5 FROM S-SS2-N 3.2 FROM C = CH2 4,5 FROM C = CH-C 4.3 Ζ C = CC2 4.0 FROM C SN 4.9 * FROM C-C 4.6 * FROM Aryl 4.3 * or 2.6 (based on benzene (delocalized electron cloud) FROM C-C = N 4.0 FROM C3-C-O- 3.6 FROM S2H-S-O- 3.8 FROM CH2-C-O- 4.1 FROM H3-C-O- 4.4 FROM C2-C = O- 3.6 FROM CH-C = O- 3.8 FROM C2-C = N ? FROM CH-C = N ? FROM S-S-I 3,1 FROM C2H-C ^ 3.3 FROM SSH-S-Ts 3.6 FROM NZ-S-S 3.8 ABOUT SLEEP 2.1 ABOUT C = O 2.1 ABOUT SOS 1.8 ABOUT \ Oh _; 1.9 * N S-MH ₂ 3,1 N C-HH-S 2.8 * N C ^ C2 2.5 * N \ Oh _; 4,5 * (maybe more in small molecules) N C n 3.2 8 C = 8 7.7 8 C-8-C ? 8 S-8-N 5,0

* Calculated based on molecular polarizabilities.

? Values can be determined based on the corresponding molecular data.

Module 1: Generating software code for calculating simulated receptors.

Step 1. Enter the parameters of the object code (command generation): ι) length (number of characters) of the code; and ΐΐ) frequency of commands.

Step 2. Initialize an empty line to write code in memory.

Operation 3. Generate a random number.

Operation 4. Based on a random number and frequency of commands, select the symbol {'0', '1', ..., '6'} to concatenate with a line of code. Repeat step 4 until the line is equal to the predefined code length.

Step 5. Print the code.

Module 2: Code Translation for Calculating Simulated Receptors.

Operation 1. Enter the initial coordinates of the polymers that make up the (constituent) receptor.

Step 2. Enter the polymer code.

Step 3. Read the first character of the code.

Operation 4. If the symbol is a rotation command, use the translation algorithm to determine the coordinates of the subunit, otherwise go to step 7.

Step 5. Enter the subunit coordinates into memory. Assign the charge 0 to the subunit.

Step 6. If the character is not the last character of the code, repeat step 3, otherwise go to the next step.

Operation 7. If the symbol is a charge command, use the translation algorithm to determine the coordinates of the subunit that do not involve rotation.

Step 8. Store the subunit coordinates in memory. Assign a charge of +1 or -1 to the subunit, depending on the character.

Step 9. If the character is not the last character of the code, repeat step 3, otherwise go to the next.

Step 1 0. Repeat steps 2 through 9 for each of the polymers that make up the receptor.

Operation 11. Print the coordinates and values of the charges of the subunit.

Module 3: Presentation of the target.

Operation 1. Enter the coordinates and radii of the target atoms (χΈ, уЕ, ζΐ _£ , radius ^ (1 = number of atoms in the target)

Enter the coordinates of the receptor (xg ,, y!}, Ζη, nucleus)

() = number of subunits in the receptor)

Step 2. Set the random values of the angles (Δθ, Δφ) and shift (k _x , k _y ).

Operation 3. Rotate and shift the coordinates of atoms by a random value.

Operation 3 a. Translate the coordinates of the target into the polar form (х ± b уГ, ζΪ £, radius ^ (θι, φι, ρι, radius))

Operation 3b. Add arbitrary changes to the angles (θ), f), P), a radius)) (θ) + Δθ) , f + Δf ^ ^ e _r radius))

Operation 3 s. Turn in Cartesian (rectangular) coordinate (θ) + Δθ), f + Δf ^ ^ e _r radius)) (x _{s x} y, ζ _χ, radius)

Operation 3ά. Add a random shift (χη _χ , _y ζη _χ , radius)) (x _x + k _x , y + k _y , ζ _χ , radius)

Step 4. Place the center of the target at the origin (0,0,0).

Operation 4a. Find the maximum and minimum values for xn), yn), ζη).

Operation 4b. Find the geometric center of the receptor ^{at the 'U} Center (IM-maksimumAI ^{minimumH-2, (ζη}

Maximum ^ζη minimum ^{) / 2.}

Operation 4s. Calculate the coordinates with respect to the center (xps ,, oops ,, ζησ ^^ ί-χη ^^ ρ, yn ₁ -Uncontrol, ζη _χ - ^ I-maximum).

Operation 5. Use the radii of the atoms and the transformed coordinates (xps), oops), ζη ^, radius)) to construct the collision surface of the target β (χ _§ ^ _§ ) = ζ _§

Operation 5a. Create a grid with an interval equal to the diameter of the receptor subunit (= 1). Coordinates of the grid: x ₈ e {1n ± (xn _mi „ _imum - xx ^ center), 1nCxy _{1 Imum} - x & _ntr ) + 1. . .0,. . 1n1 (xn _maximum - xn center) * E 1n £ (xn _maximum * ^^ center)} y ₈ e {1n1 (yn _minimum - yn ^ lLnn (y ^ at _least - _ynntr ) + 1 _... 0.,. . 1n1 (yn maximum * Uaenter) - E b * (Yn maximum * Y ^ center)}

Set zero initial values of ё (х§, у§) at all points of the coordinate grid.

Operation 5b. For each atom (ΐ), set the value g (x _§ , y _§ ) (weight) of each grid point (x _§ , y _§ ) in accordance with the following rule:

For ί = from 1 to the number of atoms in the target

If (xps _r x _p ) ² + (ynC) -ur) ² <radius) ² , then e (x§, y§) = minimum (§ (x§, y _§ ), ζη ^ is the radius))

Otherwise

If (xps _r x _p ) ² + (ynC) -ur) ² <radius) + .5) ² , then § (x§, y _§ ) = minimum (§ (x _§ , y _§ ), ζη ^ is the radius ^s ! ^{/ 2))}

Also e ^(x § ^, y§ ⁾ = ^minim y ^m (e ^ yD ⁰⁾

Then ί

Step 6. Place the center of the receptor at the origin (0,0).

Operation 6a. Find the maximum and minimum values of xm ,, γη and ζη.

Operation 6b. Find the geometric ^{center of the receptor xg} center = ^{( xg} maximum ^-xG minimum ^{) / 2} , ^yr center = ^{( yg} maximum ^-yG minimum) / 2, ^ ^G center ⁽ ^ ^G maximum “^ ^G minimum ^{) / 2}

0, .0,

Operation 6s. Calculate the coordinates of the receptor relative to the center (xc, yc ^ ^ ζο _ί) = (χΓ _ί ^x center, ^y n ^y Center, ^G ^ 1 ^- ^ ^max) ·

Operation 7. Build the collision surface of the receptor s (x ₈ , y ₈ ) = 7 ₈ using the coordinates of the receptor relative to the center in accordance with the following rule:

Set all the initial values of s (xs ₈ , us ₈ ) to

For = = from 1 to the number of subunits in the receptor, if ζ ^> 8 (xc |, usr, then 8 (χο ^ θ |) = ζθί

Go to the next й.

Step 8. Find the minimum distance between the collision surface of the receptor and the collision surface of the target. Calculate the difference matrix 4 (x _§ , y _§ ) as follows for all x ₈ e {1n £ (xn _minimum - xd, _e11tr ), 1n £ (xx wxhm Xntr) + 1

1P £ (x _max - x center) 1, 1P £ (x _max - x _TsS11tr)} and {In _AU 1P £ (y _minimum - the UOC _UNT p), Kup ™ ,, ^ - yu, _Centralized) + 1 1P £ (yn _maximum - _ynpt ) - 1, 1n £ (yn _maxnum - _ynpt )}

Calculate 4 (x _§ , y _§ ) = (H (x _§ , y _§ ) - ζη _m minimum + ^ζп maximum ⁾ + ( ⁸ ( ^x §, у §) + ^ minimum ^- ^ maximum)

For all x _§ , y _§, find the minimum value 4 (x _§ , y _§ ) = 4 _mb c1 _{m | n} means the minimum distance.

Step 9. Transform the coordinates of the collision configuration (collision) of the target and receptor.

For the receptor:

(hretseptorr uretseptorr ζretseptor) = (^ xc, ^yc p ^ζ ^ + Minimum ^- ^ max)

For a target _(s hmishen umishen ζmishen _{s 1)} = (hηs _{1 x} ups, ζη ^ - + ζpminimum

2 ' ⁿ max-ym1n ⁾ .

Operation 10 Use _(s hmishen umishen ζmishen _{s ^)} and (hretseptorr uretseptorr ζretseptor ^) for the affinity calculations.

Repeat steps 2–9 to calculate the configuration of each test target.

Module 4: Calculating Affinity.

Operation 1. Type coordinates collision target and receptor (hmishen umishen ^ _{s ^} ζmishen) and (hretseptorr uretseptorr ζretseptor ^) where ι = number of atoms in the target, d = number of subunits in the receptor.

Step 2. Enter the dipole moment values for the target ((ρ (ΐ)) dip (1)

Enter the charge values for the receptor charge 0)

Step 3. Enter the threshold value for calculating proximality ΤΗΚΕ8ΗΟΡΌ (threshold)

Step 4. Calculate the dipole affinity

Operation 4a. For each subunit charged (charge 0) = 0) to calculate e (1, j) = dip (1) / _(r hmishen hretseptor ^ ² + (umishen - uretseptorr ^ ² + target retseptorr ^ ^{2) 1.5.}

Operation 4b. Calculate the sum e (0 for all combinations of v and d with a charge of 0) ^ 0.

(DIPOLE) EYROBE = Her (C)

Operation 5. Calculate the proximal value (this operation can be replaced by a calculation based on polarizability)

Operation 5a. For each atom of the target with No. (1) | <0.75

Calculate 1 (C) = ((target _г receptor ² + (target - ureceptor ² + ^ target ζ receptor ^) ² ) ^0.5

If 1 (0 <THRESHOLD, then prox (0 = 1

Operation 5b. Calculate the sum rgox (0 for all combinations of v and th with | dip (s) | <0.75

ΡΚΟΧΙΜΙΤΥ (PROXIMALITY) = Σ rgo ^{x (} O

Step 6. Calculate the affinity for the target-substrate combination = affinity (ΑΡΡΙΝΙΤΥ)

Affinity = (PROXIMALITY / th) (PROXIMILITY / 10000 + DIPOLE),

ΑΡΡΙΝΙΤ = (ΡΚΟΧΙΜΙΤΥ / d) (ΡΚΟΧΙΜΙΤΥ / 10000) + ΌΙΡΟΡΕ)

Module 5: Calculating the degree of compliance.

Operation 1. Enter the efficacy or affinity values of a known target (y _k ), k = number of tested targets

Step 2. Enter the coordinates of the collision (collision) of the targets and the receptor (target _A , target _A , ζ target _^ ) and (chreceptor ureceptor ζ receptor ^), where 1 _k = number of atoms in the target k, и = number of subunits in the receptor

Step 3. Enter the number of target orientations to be tested (= t).

Step 4. Use Module 5 to obtain affinity values for each target and target orientation (^ ΑΡΡΙΝΙΤΥ ^).

Step 5. Determine the maximum affinity (MS _k ) (MA _k ) and total affinity (SS _k ) (8A _k ) for each target.

Operation 6. Calculate the correlation coefficients g _MS ² (gMA ² ) for maximum affinity (MSC; MA _k ) relative to the efficiencies of a known target or the value of affinity characteristics (y _k ) and g _8A ² for total affinity (SS _k ; 8A _k ) relative values of efficiency and affinity of a known target (uk).

Operation 7. Calculate the compliance coefficient Ρ ^Ρ = ^(g MA ^{2 X} G _§ a ^{2) 0.5}

Or

Operation 6 '. Calculate the correlation coefficients g _MA ² for maximum affinity (MAk) relative to the affinity efficiency values of a known target (yk) and g8A-MA ² for total affinity (8Ak) - maximum affinity relative to the efficiency and affinity values of a known target (yk).

Operation 7 '. Calculate the compliance coefficient Ρ ^Ρ = ^(g MA ^{2 X (1-G} 8A-MA ^{2)) 0.5}

Module 6: Generate a genotype with a minimum degree of affinity.

Operation 1. Set minimum compliance threshold

Step 2. Generate a Random Genotype (Module 1)

Step 3. Translate the genotype to build the phenotype (Module 2)

Step 4. Check the affinity of the phenotype for the targets (Modules 3, 4, 5, 6)

Step 5. If the degree of phenotype compliance exceeds the threshold value, stop generating the code and proceed to the execution of Phase 2. Otherwise, repeat steps 1-5.

Module 7: Multiple Mutation

Step 1. Enter the primary code (from Phase 1).

Step 2. Enter the number (= c |) of mutations in the code (in this case, 2.5-5% of the characters in the genotype mutate).

Step 3. Enter the population size (= p).

Step 4. Randomly select a position in the genotype.

Step 5. Replace the code symbol at this position with another, arbitrarily selected symbol.

Step 6. Repeat steps 4 and 5 until you reach the number q (q times).

Operation 7. Repeat operations 4-6 until the total amount of newly generated codes reaches p.

Step 8. Apply Modules 1-6 to verify the mutant population. Select the subpopulation with the highest selectivity for use in Phase 3.

Module 8: Recombination.

Operation 1. Enter the population size (= p).

Step 2. Choose two codes from a population generated by Phase 2 at random.

Step 3. Randomly select a position in the genotype.

Step 4. Create a random number for the number of characters exchanged.

Step 5. Swap the characters between the codes, starting from the selected position.

Step 6. Repeat steps 2–5 until p new genotypes are created.

Step 7. Apply Modules 2-6 to verify compliance with the mutant population. Choose the subpopulation with the highest selectivity for the next series of recombinations or for maturation in Phase 4.

Module 9: Ripening.

Operation 1. Enter the start code generated from Phase 3.

Step 2. Enter the number of iterations.

Step 3. Choose an arbitrary position in the original genotype.

Operation 4. Replace the code symbol in this position with another, arbitrarily selected symbol.

Step 5. Verify the selectivity of the precursor code (P _p ) and mutation product (P _m ) using Modules 2-6.

Step 6. If P _m > P _p , replace the original genotype with the mutation product.

Step 7. Repeat steps 3–6 for the required number of iterations.

Module 1 0: Creating carbon skeleton code. containing cycles (6 membered cycles, entry point = exit point).

Operation 1. Enter (length) the number of characters in the code

Enter ν1, ν2, ν3,., Νη (frequency of substituent groups).

Enter pgb_hfd (cycle code sequence frequency). (0 <pgb_gfd <1)

Operation 2. Initialize pptesobe =.

Initialize 8soeb-sobe =.

Initialize bygb-sobe =.

Step 3. Create character strings.

Repeat step 4 until you get the code of the specified length.

Operation 4a. If ρο г ш д>> gaibot (random number) (0 <gaibot <1) then

Assigning characters to a ring (bath condition).

Set the new_symbol_1 (name_cNagas1eg_1) to a randomly selected member from {'431413', '314134', '141343', '132132', '321321', '213213', '123123', '231231', '312312', '421412', '214124', '141242', '324234', '242343', '423432'}

Assigning characters to alternates.

Set the new_symbol_of_2 (em_syagas1eg_2) to 6 members arbitrarily selected from {c1, c2, c3,., Si} using the frequencies ν1, ν2, ν3,., Νΐ'ΐ.

(c1. c2 are symbols defining various functional groups).

Assigning symbols to valencies of substituents.

Set new_symbol_of_3 (em_syagas1eg_3) to 6 randomly selected members from {'1', '2', '3', '4'}.

Otherwise

Operation 4b. Assignment of single (non-ring) characters to the primary code.

Set user_cNagas1eg_1 (new_symbol_1) to a randomly selected member from {'1', '2', '3', '4'}.

Assigning characters to alternates.

Set em_cNagas1eg_2 to an arbitrarily selected term from {c1, c2,., Si}, using the frequencies ν 1, ν2, ν3,., Νιι.

Assigning symbols to valencies of substituents.

Set ne \\ _ c11agas1eg_3 to a randomly selected member from {'1', '2', '3', '4'}.

Operation 4s. Concatenate new characters with lines of code Ppt_soDe = Ppt_cODe apD ne \ y_sNagas1eg_ 1 (primary_code = primary_code and new_symbol_1)

ZssopD_soODs = ZssopD_soODs apD ps \\ _ s Nagas1sg_2

TYGD_SODE = TYGD_SODE apD ps \\ _ sNagas1sg_3

Module 11: Multiple Point Mutation

Operation 1. Enter the primary code.

Step 2. Set the number (= 4) of mutations per code (In this case, 2.5-5% of genotype characters mutate).

Step 3. Enter the population size (= p).

Step 4. Randomly select a position in the genotype.

Step 5. Replace the code characters at this position in each code vector with other, arbitrarily selected characters.

Step 6. Repeat steps 4 and 5 to c | time.

Step 7. Repeat steps 4–6 to generate a total of not new codes.

Step 8. Verify that each member of the mutant population matches.

Select the most suitable subpopulation for use in recombination or additional multiple mutation 3.

Module 1 2: Mutation Sequences.

Operation 1. Set P _PE L, p _U y, and P Ruse. as threshold values for the presence of a mutation (0 <P _x <1).

Operation 2. Generate an arbitrary position (= x) in the code (0 <P <Code length).

Operation 3. Generate a random sequence length (= b) (0 <b <code length - x).

Operation 4. Copy the sequence from the code, starting with x and with the extension to the total amount of b characters.

Operation 5. If 0 <P _uu <Random number <1 Then

Reverse the order of the characters in the string.

Operation 6. If 0 <P _CuP <Random Number <1 Then

Copy the sequence and concatenate the copy to the sequence.

Operation 7. If 0 <P _CEB <Random Number <1 Then

Eliminate b characters from the code, starting at position x.

Otherwise

Replace the sequence in the code with the sequence generated in steps 5 and 6.

Step 8: If 0 <P _SW <random number <1 Then

Generate a position (= y) in the code arbitrarily (0 <y <code length).

Insert the sequence generated in steps 5 and 6 into position y.

Module 13: Recombination.

Step 1. Set the population size (= p).

Step 2. Randomly select two codes from the population generated by the multiple mutation.

Step 3. Choose an arbitrary position in the genotype.

Operation 4. Generate a random number for the number of characters to be exchanged.

Step 5. Swap the characters between each of the three code vectors, starting from the selected position.

Step 6. Repeat steps 2–5 until p new genotypes are generated.

Step 7. Check the correspondence of each ligand in the resulting mutant population. Select the most appropriate subpopulation for the next series of recombination or for maturation.

Module 1 4: Ripening.

Step 1. Enter the initial code formed from recombination.

Step 2. Set the number of iterations.

Step 3. Specify an arbitrary position in the original genotype.

Step 4. Replace the code characters at these positions in each code vector with other, arbitrarily selected characters.

Step 5. Verify that the precursor code (P _p ) and mutation product (P _m ) match using Modules 4 and 5.

Step 6. If P _m > P _p , replace the precursor genotype with the mutation product.

Step 7. Repeat steps 3-6 until the desired number of iterations.

Claims (52)

one . A method for computer-aided design of chemical structures having a pre-selected functional characteristic, comprising the following stages: (a) creating a physical model of a phenotype of a simulated receptor encoded as a linear sequence of characters, and creating a set of target molecules having at least one common quantifiable functional characteristic; (b) for each target molecule (ί) calculating the affinity between the receptor and the target molecule in each of the multiple orientations using the calculation of the effective affinity; (ίί) calculating the total affinity by summing the calculated affinity values; (iii) determination of maximum affinity; (c) applying the calculated values of the total and maximum affinity for (ί) calculating the correlation coefficient of the maximum affinity between the maximum affinity and a quantifiable functional characteristic; (ίί) calculating the correlation coefficient of the total affinity between the total affinity and a quantitatively determined functional characteristic; (d) the application of the maximum correlation coefficient and the total correlation coefficient to calculate the correspondence coefficient; (e) changing the structure of the receptor and repeating steps (b) to (d) to obtain a population of receptors having a pre-selected matching coefficient; (e) creating a physical model of a chemical structure encoded as a molecular linear sequence of characters; calculating the affinity between the chemical structure and each receptor from a variety of orientations using the indicated method for calculating the effective affinity, using the calculated affinity values to calculate the degree of affinity correspondence; (g) changing the chemical structure to obtain a variant of the chemical structure and repeating stage (e); and (h) preservation and further change of those variants of the chemical structure, the degree of affinity of which approaches a predetermined (selected) degree of affinity. 2. The method according to claim 1, characterized in that the linear sequence of characters encoding the indicated phenotype of the receptor is obtained by generating a linear sequence of characters of the receptor, which encodes the spatial distribution and charge, and also characterized in that the step of creating a physical model of the chemical structure includes generating said a molecular linear sequence of characters that encodes a spatial distribution and charge. 3. The method according to claim 2, characterized in that said calculation of effective affinity includes two criteria, the first being a proximal criterion, characterized in that the uncharged portions of the indicated simulated receptor are sufficiently close to the non-polar portions of the indicated molecular structure so that the intensity of the effective London dispersion becomes it is possible to determine, and the second is the total intensity of the electrostatic charge-dipole interaction arising between the charged sections of the indicated m receptor and to model the dipoles present in the said molecular structure. 4. The method according to claim 2, characterized in that said step of calculating the degree of affinity correspondence includes calculating the total and maximum affinity between the molecular structure and each receptor, and the degree of correspondence is calculated as follows: Σ {| calculated maximum affinity - maximum affinity of the target | / maximum affinity of the target}, and also characterized in that said pre-selected degree of compliance is practically zero. 5. The method according to claim 2, characterized in that said step of calculating the degree of affinity correspondence includes calculating the total and maximum affinity between the molecular structure and each receptor, and the degree of correspondence is calculated as follows: Σ {(| estimated maximum affinity is the maximum affinity of the target | / 2 x maximum affinity of the target) + (| calculated total affinity is the total affinity of the target |)}, and also characterized in that the pre-selected degree of compliance is practically zero. 6. The method according to claim 2, characterized in that said correlation coefficient of the total affinity is g _ZA ² , said correlation coefficient of maximum affinity is gMA ² , and also that said correspondence coefficient means P = (gMA ² x g _ZA ² ) ^0.5 , as well as the fact that the pre-selected matching coefficient is practically equal to one. 7. The method according to claim 2, characterized in that said correlation coefficient of total affinity is g _ZA-MA ² , said correlation coefficient of maximum affinity is gMA ² , also characterized in that said correspondence coefficient is P = (gMA ² x (1 ^g s _A- m _A ² )) ⁰ ' ⁵ , and also by the fact that said pre-selected coefficient is practically equal to unity. 8. The method according to claim 2, characterized in that said molecular linear symbol sequences include a plurality of symbol triple sequences, the first symbol of said triple being arbitrarily selected from the set of first symbols determining the position and identity of the populating atom in the molecular skeleton of said molecular structure, the second symbol the specified triplet is arbitrarily selected from the set of second symbols that determine the identity of the substituent associated with the specified populating atom, and the third symbol seemed triplet randomly selected from a set of symbols tret61 determining the location of said substituent at precisely defined by the first character of the triplet. 9. The method according to claim 8, characterized in that the molecular linear sequence of characters is decoded using an efficient molecule arrangement algorithm that sequentially moves each triplet of the specified molecular linear sequence and then saturates the unfilled positions at the specified molecular skeleton with hydrogen atoms. 10. The method according to claim 9, characterized in that the stage of changing the molecular structure includes at least one of the following operations: ί) mutation of the indicated genotype of the molecule through arbitrary mutual exchange of at least one of the first, second or third a symbol with at least one triplet from the sets of corresponding symbols, ϊϊ) a deletion in which the triplet is removed from the genotype of the molecule, ίίί) duplication, in which the triplet in the genotype of the molecule is duplicated, ίν) inversion, in which the sequential order of one or more triplets in the genotype of the molecule are reversed, and ν) the insertion, in which a triple from the genotype of the molecule is introduced into a different position in the genotype of the molecule. 11. The method according to p. 1 0, characterized in that the stage of mutation of the indicated molecular genotypes includes the recombination of randomly selected pairs of these stored mutant molecular genotypes, while the corresponding characters in the indicated molecular linear sequences are mutually exchanged. 12. The method according to claim 2, characterized in that each symbol in a linear sequence of receptor symbols defines either a spatial rotation command or a charged site without rotation. 1 3. The method according to p. 1 2, characterized in that said receptor phenotype contains at least one linear polymer with a plurality of subunits, wherein one of said subunits is the first subunit in said at least one linear polymer. 14. The method of claim 13, wherein said linear sequence of receptor symbols is decoded using an efficient receptor assembly algorithm in which rotation instructions applied to each subunit following said first subunit are relative to the position of said first subunit. 15. The method according to p. 14, characterized in that said symbols define a program of spatial rotation commands for no rotation, right rotation, left rotation, upward rotation, downward rotation, and the symbols determine charge site code for a positively charged site without rotation and negatively charged site without turning. 16. The method according to 14, characterized in that said subunits are practically spherical with Van der Waals radii almost equal to the van der Waals radius of hydrogen. 17. The method of claim 15, wherein the step of changing said receptor genotype includes at least one of the following operations: ί) a deletion in which a symbol is deleted from the receptor genotype, ίί) a duplication in which the symbol in the genotype is duplicated receptor, ίίί) inversion, in which the sequential order of one or more characters in the receptor genotype is reversed, and ίν) insertion, in which the symbol from the receptor genotype is introduced into a different position in the genotype. 18. The method according to 17, characterized in that the stage of mutation of the indicated genotypes of the receptors includes the recombination of randomly selected pairs of the indicated stored genotypes of the mutant receptors, while the corresponding characters in the indicated linear sequences of the receptors are mutually exchanged. 19. A method of computer screening of chemical structures with pre-selected functional characteristics, which consists in (a) creating a genotype of a simulated receptor by generating a receptor linear sequence of characters that encodes a spatial distribution and charge; (b) decoding the genotype in order to create a receptor phenotype comprising at least one target molecule exhibiting a selected functional characteristic, calculating the affinity between the receptor and each target molecule in a variety of orientations using effective affinity calculation, calculating the total and maximum affinity between each target molecule and receptor, calculating the total affinity correlation coefficient for total affinity relative to the specified functional characteristics of the target molecule the correlation coefficient and the maximum affinity for maximum affinity relative to the specified functional characteristics (targets), and the calculation of the coefficient of compliance, depending on the specified correlation coefficients of the total and maximum affinity; (c) mutation of the receptor genotype and repetition of step (b) and conservation and mutation of receptors with increased compliance coefficients until receptor populations with pre-selected matching coefficients are obtained; further (d) calculating the affinity between the screened chemical structure and each receptor in a variety of orientations, using the indicated calculation of effective affinity, calculating the degree of affinity correspondence, which includes calculating the total and maximum affinity between the compound and each receptor and comparing at least one of the values of the specified total and maximum affinity with the values of the total and maximum affinity between the specified at least one target and the specified population of receptors, etc. and this comparison is an indicator of the degree of functional activity of the indicated chemical structure relative to the specified at least one target molecule. 20. The method according to p. 19, characterized in that said calculation of effective affinity contains two criteria, the first being a proximal criterion in which uncharged portions of these simulated receptors are close enough to non-polar portions of the indicated molecular structure to make it possible to determine (intensity) effective Of the London dispersion, and the second is the total force of the electrostatic charge-dipole interaction arising between the charged sections of the specified modeled etseptora and dipoles existing in said molecular structure. 21. The method according to claim 20, characterized in that the degree of compliance is calculated as Σ {| estimated maximum affinity - maximum target affinity | / maximum affinity of the target}. 22. The method according to claim 20, characterized in that the degree of compliance is calculated as Σ {(| estimated maximum affinity - maximum affinity of the target | / 2 x maximum affinity of the target) + (| calculated total affinity total affinity of the target | / 2 x total affinity targets)}. 23. The method according to claim 20, characterized in that the correlation coefficient of the total affinity is t _FOR ² , the specified correlation coefficient of the maximum affinity is gMA ² , and that the specified compliance coefficient is P = (gMA ² x g _FOR ² ) ⁰ · ⁵ , as well as the fact that the pre-selected matching coefficient is practically equal to unity. 24. The method according to claim 20, characterized in that said correlation coefficient of total affinity is t _ZA-MA ² , said correlation coefficient of maximum affinity is gMA ² , and that said correspondence coefficient is P = (gMA ² x (1- g _ZA-MA ² )) ⁰ · ⁵ , as well as the fact that the specified pre-selected matching coefficient is practically equal to one. 25. The method according to claim 20, characterized in that each symbol in the receptor linear sequence of characters determines either a rotation command in space or a charged site without rotation. 26. The method according A.25, characterized in that the phenotype of the receptor contains at least one linear polymer with many subunits, and one of these subunits is the first subunit in the specified at least one linear polymer. 27. The method according to p. 26, characterized in that the indicated linear receptor sequence of characters is decoded using an efficient receptor assembly algorithm in which (the algorithm) rotation commands applied to each subunit following the specified first subunit are set relative to the initial position of the specified first subunits. 28. The method according to item 27, wherein said symbols define a program of rotation commands in space for no rotation, right turn, left turn, turn up, turn down, and also that symbols determine the code of charge sites for a positively charged site without turning and negatively charged site without turning. 29. The method of claim 28, wherein said subunits are substantially spherical with van der Waals radii substantially equal to the van der Waals radius of hydrogen. 30. The method according to item 27, wherein the stage of mutation of the indicated receptor genotype includes at least one of the following operations: ί) deletion, which removes the symbol from the receptor genotype, ίί) duplication, in which the symbol in the genotype is duplicated receptor, ίίί) inversion, in which the sequential order of one or more characters in the receptor genotype is reversed, and ίν) insertion, in which the symbol from the receptor genotype is introduced into a different position in the genotype. 31. The method according to p. 30, characterized in that the stage of mutation of the indicated genotypes of the receptors includes the recombination of randomly selected pairs of these stored genotypes of the mutant receptors, while the corresponding characters in the indicated linear sequences of the receptors are mutually exchanged. 32. A method for the computer construction of simulated receptors that mimic biological receptors exhibiting selective affinity for compounds with similar functional characteristics, comprising the steps of: (a) creating a simulated receptor genotype by generating a receptor linear sequence of characters that encodes a spatial distribution and charge; (b) decoding the genotype in order to create a receptor phenotype, create a set of target molecules having common similar functional characteristics, calculate the affinity between the receptor and each target molecule in many orientations using the calculation of effective affinity, calculate the total and maximum affinity between each target molecule and receptor , calculation of the correlation coefficient of total affinity relative to the functional characteristics of each target molecule and the correlation coefficient of maximum affinity from ositelno said functional characteristics for each target molecule and calculation of the corresponding coefficients, dependent on said correlation coefficients and maximum cumulative affinity for each target molecule; and (c) mutations of the genotype and repetition of step (b) and conservation and mutation of receptors with increased compliance coefficients until a population of receptors with pre-selected matching coefficients is created. 33. The method according to p, characterized in that each symbol in the receptor linear sequence of characters determines either a rotation command in space or a charged site without rotation. 34. The method according to p. 33, characterized in that the receptor phenotype contains many linear polymers with many subunits, each linear polymer being encoded by a corresponding linear sequence of characters and one of these subunits is the first subunit in the specified at least one linear polymer . 35. The method according to clause 34, wherein said linear receptor linear sequence of characters is decoded using an efficient receptor assembly algorithm in which (algorithm) rotation commands applied to each subunit following said first subunit are set relative to the starting position of said first subunits. 36. The method according to claim 35, characterized in that said symbols specify a program of rotation commands in space for no rotation, right rotation, left rotation, up rotation, downward rotation, and that symbols define a charge site code for a positively charged site without turning and negatively charged site without turning. 37. The method according to clause 36, wherein said subunits are almost spherical with van der Waals radii practically equal to the van der Waals radius of hydrogen. 38. The method according to clause 35, wherein the stage of mutation of the indicated receptor genotype includes at least one of the following operations: ί) deletion, in which the symbol is removed from the receptor genotype, ίί) duplication, in which duplicates in the receptor genotype a symbol, ίίί) an inversion in which the sequential order of one or more characters in the receptor genotype is reversed, and ΐν) an insertion in which a symbol from the receptor genotype is introduced into a different position in the genotype. 39. The method according to § 38, characterized in that the stage of mutation of the genotypes of the indicated receptor includes the recombination of randomly selected pairs of these stored genotypes of mutant receptors, while the corresponding characters in the indicated linear receptor sequences are mutually exchanged. 40. The method according to p. 33, characterized in that said calculation of effective affinity contains two criteria, the first being a proximal criterion in which uncharged portions of these simulated receptors are close enough to non-polar portions of the indicated molecular structure so that the intensity of the effective London dispersion can be determined and the second is the total intensity of the electrostatic interaction of the charge-dipole arising between the charged sections of the specified receptor and dipoles present in the indicated molecular structure. 41. The method according to p. 40, characterized in that the correlation coefficient of the total affinity is g _FOR ² , the correlation coefficient of the maximum affinity is gMA ² , and also characterized in that the coefficient of compliance means P = (gMA ² x g _FOR ² ) ⁰ · ⁵ , as well as the fact that the pre-selected matching coefficient is practically equal to one. 42. The method according to claim 40, wherein said total affinity correlation coefficient is g _ZA-MA ² , said maximum affinity correlation coefficient is gMA ² , also characterized in that said correspondence coefficient is P = (gMA ² x (1g _8A ² )) ⁰ ' ⁵ , and also by the fact that said pre-selected coefficient is practically equal to unity. 43. A method for the computer creation of chemical structures with pre-selected functional features, comprising the following stages: (a) creating a physical model of a receptor and a set of target molecules, wherein the target molecules have at least one general quantifiable functional characteristic; (b) for each target molecule (ί) calculating the affinity between the receptor and the target molecule in each of the multiple orientations using the calculation of the effective affinity; (ΐΐ) calculating the total affinity by summing the calculated affinity values; (iii) determination of maximum affinity; (c) applying the calculated values of the total and maximum affinity for (ί) calculating the correlation coefficient of the maximum affinity between the maximum affinity values and a quantitatively determined functional characteristic; (i) calculating the correlation coefficient of the total affinity between the values of the total affinity and a quantitatively determined functional characteristic; (d) applying the maximum correlation coefficient and the total correlation coefficient to the calculation of the correspondence coefficient; (e) changing the structure of the receptor and repeating steps (b) to (d) to obtain a receptor population with a pre-selected matching coefficient; (f) creating a physical model of the chemical structure, calculating the affinity between the chemical structure and each receptor in a variety of orientations using the specified calculation of effective affinity, using calculated affinity values to calculate the degree of affinity correspondence; (g) changing the chemical structure in order to obtain a variant of the chemical structure and repeating stage (e); and (h) preservation and further change of those variants of the chemical structure, the degree of affinity of which approaches a predetermined degree of affinity. 44. The method according to item 43, wherein the step of creating a physical model of the receptor includes generating a linear receptor sequence of characters that encodes a spatial distribution and charge, and the fact that the step of creating a physical model of the chemical structure includes generating a linear sequence of characters that encodes a spatial distribution and charge. 45. The method according to claim 44, wherein said linear symbol sequences for said chemical structure comprise a plurality of consecutive symbol triplets, wherein the first symbol of said triplet is arbitrarily selected from the set of first symbols determining the position and identity of the populating atom in the skeleton of the molecule of said chemical structure , the second symbol of the specified triplet is arbitrarily selected from the set of second symbols that determine the identity of the substituent associated with the specified populating atom, and the third symbol of the indicated triplet is arbitrarily selected from the set of third symbols that determine the location of the substituent at the atom, precisely determined by the first symbol of the triplet. 46. The method according to item 45, wherein the linear sequence of characters of the chemical structure is decoded using an efficient molecule assembly algorithm that sequentially moves each triplet from the specified molecular linear sequence, and then saturates the unfilled positions at the specified molecular skeleton with hydrogen atoms. 47. A method of encoding a chemical structure containing atomic elements, the method comprising creating a linear sequence of characters that encodes the spatial distribution, relative position of an atom, type of bond, and charge for each atom in order to determine a single three-dimensional conformation of said chemical structure. 48. The method according to clause 47, wherein said linear sequence of characters for the specified chemical structure contains many consecutive triplets of characters, and the first character of the specified triplet is arbitrarily selected from the set of first characters that determine the position and identity of the populating atom in the skeleton of the molecule of the specified chemical structure , the second symbol of the specified triplet is arbitrarily selected from the set of second symbols that determine the identity of the substituent associated with the specified populating atom, and the third symbol of the indicated triplet is arbitrarily selected from the set of third symbols that determine the location of the substituent at the atom, precisely determined by the first symbol of the triplet. 49. The method according to item 45, wherein the linear sequence of characters is decoded using an efficient molecule assembly algorithm that sequentially moves each triplet from the specified linear sequence of characters and then saturates the unfilled positions of this molecular skeleton with pre-selected atoms. 50. The method according to 49, including the operation of storing the specified linear sequence of characters in a computer storage device. 51. The method according to claim 19, characterized in that said functional characteristic is biological toxicity. 52. The method according to claim 19, characterized in that said functional characteristic is catalytic activity.