[go: up one dir, main page]

WO2014012309A1 - 一种分子体积计算方法及两分子的形状比较方法 - Google Patents

一种分子体积计算方法及两分子的形状比较方法 Download PDF

Info

Publication number
WO2014012309A1
WO2014012309A1 PCT/CN2012/084713 CN2012084713W WO2014012309A1 WO 2014012309 A1 WO2014012309 A1 WO 2014012309A1 CN 2012084713 W CN2012084713 W CN 2012084713W WO 2014012309 A1 WO2014012309 A1 WO 2014012309A1
Authority
WO
WIPO (PCT)
Prior art keywords
molecule
volume
gaussian
intermolecular
calculating
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2012/084713
Other languages
English (en)
French (fr)
Inventor
徐峻
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sun Yat Sen University
Original Assignee
Sun Yat Sen University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sun Yat Sen University filed Critical Sun Yat Sen University
Priority to JP2014545073A priority Critical patent/JP2015502611A/ja
Priority to US14/351,978 priority patent/US9811642B2/en
Priority to EP12881232.8A priority patent/EP2752784A4/en
Publication of WO2014012309A1 publication Critical patent/WO2014012309A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/50Molecular design, e.g. of drugs
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16BBIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B15/00ICT specially adapted for analysing two-dimensional or three-dimensional molecular structures, e.g. structural or functional relations or structure alignment
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/30Prediction of properties of chemical compounds, compositions or mixtures
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/40Searching chemical structures or physicochemical data

Definitions

  • the matching and quantitative comparison of the three-dimensional shape and pharmacodynamics of two molecules is the main method in drug design.
  • the molecular volume is related to the shape, and the shape of the surface determines the physicochemical properties of the molecule.
  • the nature determines the biological activity of the molecule.
  • the atoms, and the radius of the sphere is the van der Waals radius of the atom.
  • the volume density function for each hard sphere is inside the sphere and 0 outside the sphere. Such a bulk density can be expanded from a single atomic hard sphere bulk density to the following formula (1):
  • fi(r) is the bulk density of atom i.
  • the volume of a diatomic molecule is equal to the sum of two atomic volumes minus the overlapping volume between them, namely V ⁇ VA+VB VAB.
  • V VA + VB + VC VAB VAOVBOV ⁇ BC For any molecule, its volume is also a series of "including ' f and 'exclude” alternating items, calculated by the following formula (2): V ⁇ j F(r) r
  • v is the volume of the atom
  • v ⁇ j is the atomic ⁇ and j (second order) superimposed volume
  • v i jk is the atomic i> j and k (. ⁇ order) superimposed volume, and so on.
  • N 1 order atom the summation above generally proceeds to the superposition of the N 1 order atom.
  • the main drawbacks of using the above method to calculate the molecular shape are: due to the discontinuity of the volume of the hard sphere model to the atomic coordinate derivative, the length of the meter is lengthened, and the accuracy is degraded.
  • the invention provides a molecular volume calculation method and a two-molecule shape comparison method to solve the technical problem of calculating the molecular volume time and the precision is not high in the prior art.
  • a molecular volume metering method comprising:
  • Step 11 reading three-dimensional structural information of the first molecule, the three-dimensional structural information including the first molecule
  • Step 12 Obtain a corresponding van der Waals radius according to the type of each atom in the molecule, and convert the three-dimensional structural information into a set of Gaussian spheres representing each atom in the first molecule, and the radius of each Gaussian sphere and corresponding
  • Step 13 calculating a stacking volume of each pair of Gaussian spheres, wherein the ij group of Gaussian spheres includes: i Gaussian spheres and jth Gaussian spheres, and the overlapping volume of the ijth Gaussian spheres group is; 4. Calculate the weight of each Gaussian sphere.
  • the weight of the i-th Gaussian sphere— ⁇ — is the volume of the ⁇ th Gaussian sphere, and k is a constant;
  • the first superimposed volume of the first molecule is ⁇ H.- w ; .v ?
  • the superimposed volume is the volume of the first molecule, where is the weight of the i-th Gaussian sphere, and is the weight of the j-th Gaussian sphere, which is the superimposed volume of the i-th Gaussian sphere and the j-th tallest spherical sphere, A is the first molecule
  • the value of k is ().
  • the i-th Gaussian sphere and the j-th Gaussian sphere in the molecule, the overlapping volume of the ij-th Gaussian sphere in the W-molecule calculates the weight of each Gaussian sphere of the first molecule, and the first Gaussian sphere in the first molecule the weight of
  • the stacking volume is ⁇ ; the weight of each Gaussian sphere in the second molecule is calculated, the weight of the ⁇
  • V is the volume of the ⁇ th ball in the second molecule
  • * ' is a constant
  • Calculate the volume of the ⁇ 3 ⁇ 4 body of the second molecule is ⁇ « ⁇ ,,' , ⁇ indicates the second a collection of all 3 ⁇ 4 s balls of molecules;
  • Step 32 Calculating the intermolecular volume of the first molecule and the second molecule under various overlapping conditions, and selecting the maximum value as the maximum intermolecular volume;
  • Step 33 Calculate the similarity between the first molecule and the second molecule according to the volume of the first molecule, the volume of the second molecule, and the maximum intermolecular volume, and compare the similarity to the comparison of the two molecular shapes. Take a look:
  • the intermolecular volume of the step 32 is an intermolecular superimposed volume, and specifically includes: calculating an intermolecular overlapping volume of the first molecule and the second molecule in a plurality of overlapping cases as an intermolecular volume, wherein 1 " For the overlapping volume of the i-th atom in the first molecule and the j-th atom in the second molecule, the maximum value is selected as the maximum intermolecular volume; the step 33 specifically includes: calculating the first molecule and the second molecule Similarity: w > with similarity as two improving:
  • the intermolecular volume of the step 32 is a comprehensive superimposed volume of the molecules of the first molecule and the second molecule, and specifically includes:
  • Step 5.1 Find the pharmacophore in each molecule and its location:
  • Step 52 determining the type of each pharmacophore in each molecule
  • Step 53 Calculating the self-integrated superimposed volume ⁇ / of the first molecule, where is the overlapping volume between the i-th pharmacophore and the j-th pharmacophore in the first molecule, and the sum of the pairs Limited to similar pharmacophores;
  • Step 3 ⁇ 4 calculating a composite volume of the second molecule, ⁇ ⁇ X ', where ' is the overlapping volume between the ninth pharmacophore and the jth pharmacophore in the second molecule, and 'The summation is limited to between similar pharmacophores;
  • Steps Calculating the intermolecular combined superposition volume of the first molecule and the second molecule 3 ⁇ 4 ⁇ * > where is the ninth pharmacophore in the first molecule and the jth pharmacophore in the second molecule The volume of the overlap, and the summation of F ti " is limited to between the same pharmacophores; Step 56, calculating the intermolecular integrated superimposed volume of the first molecule and the second molecule in various overlapping cases as an intermolecular Volume, the maximum value of 0 12 ⁇ is selected as the maximum intermolecular volume; the step 33 specifically includes: calculating the similarity between the first molecule and the second molecule - Gi ⁇ , with similarity. As two molecular shapes Comparing results.
  • the invention reduces the computational complexity of the quantitative comparison of the three-dimensional shape of the molecule, and improves the accuracy of the calculation It can be applied to the mathematical expression of molecular shape, the comparison of drug intermolecular shapes, and the comparison of drug molecular effects. These comparison methods use virtual screening of drug molecules.
  • ⁇ i is a schematic diagram of the calculation of the molecular overlap volume.
  • Figure 2A is a comparison of the calculated molecular volume ((V) with the value obtained by the hard sphere model (HSV)) in the present invention.
  • ⁇ 3 is the calculated molecular volume flow circle of the present invention.
  • Figure 4 is a flow circle for calculating the maximum overlap of molecules and comparing the similarity of shapes of the present invention.
  • Figure ⁇ is a flow country for comparing the two molecular shapes and the similarity of pharmacophores in the present invention.
  • Figure 6 is a process diagram of a set of molecules superimposed in the present invention.
  • the present invention provides a method for quantitative comparison of three-dimensional shapes of molecules, which reduces the computational complexity of quantitative comparison of molecular three-dimensional shapes and improves calculation accuracy.
  • the Gaussian function is used instead of the hard sphere density function f&d, while the rest of equations (i) and (2) remain unchanged.
  • the Gaussian bulk density of each atom in the molecule gr ⁇ (3).
  • r is a coordinate of any point
  • .r f is the coordinate of the atom
  • ⁇ ' is the van der Waals radius of the atom
  • is a tunable parameter, usually taking a value of 2.7. Since the Gaussian function is easy to find the integral and the derivative, the computational complexity is reduced. However, when the molecular shape is calculated using the formula (2), the sum of the high-order terms is combined. i, therefore, the formula (: ⁇ is approximated as follows, that is, the molecular bulk density is reduced to the atomic volume
  • Gong Gong (4) calculates the molecule itself Between 2-6. Since the calculation of molecular shape similarity involves the intermolecular and overlapping of the molecules themselves
  • the van der Waals radius of the atom, and the tunable parameter in this radiant value is 2 ⁇ , such a value
  • the stacking volume is equal to the atomic volume itself, ie 4? ⁇ ⁇ /3.
  • the calculated molecular il body is superimposed on the volume ⁇ several times its correct value.
  • the weighting factor is an empirical function of the atom in the molecular VU context. This empirical function satisfies the following conditions: The weight should be h. If it overlaps with other atoms, its value is less than ⁇ ., and the larger the overlap, the smaller the weight.
  • the present invention proposes the following method to determine the weight - one ⁇ -, formula ( 7 ) where is Universal experience parameters, which can be obtained by fitting: k
  • becomes the overlapping volume of the molecule itself, which should be equal to the volume of the molecule itself.
  • the k constant in the formula above is the volume and molecule of the molecule itself.
  • Fig. 2A shows the relationship between the molecular volume and the actual volume of the molecule obtained by the formula (8), and the correlation between the two approaches i.
  • Figure 2B shows the actual superimposed volume of the molecular volume obtained by the simple addition method in the literature.
  • the correlation is much lower than the correlation obtained by the present invention in the cabinet 2A.
  • the present invention improves the accuracy of the intermolecular overlap volume, the accuracy of calculating the shape similarity between molecules by the formula (5) is also improved.
  • the weight can also be used as a test parameter according to the atomicity of the atom in the molecule (atomic hybrid state, bond type and number, adjacent atom type, etc.) to determine the weight and molecular conformation Relationship, such a characteristic ⁇ 3 ⁇ 4 ⁇ reduces the complexity of the flexible overlap between molecules and the flexibility of molecular shape similarity
  • ⁇ 3 ⁇ 4 ⁇ reduces the complexity of the flexible overlap between molecules and the flexibility of molecular shape similarity
  • Step S301 reading three-dimensional structure information of the first molecule, the structure information includes a first sub-step S302, obtaining a corresponding van der Waals radius according to the type of each atom in the first molecule, and converting the three-dimensional structure information into a group A Gaussian sphere representing each of the 4th atoms, the radius of each Gaussian sphere is the same as the van der Waals radius of the corresponding atom, and the position of each Gaussian sphere is the same as the coordinates of the corresponding atom, and the specific expression of the Gaussian function is obtained by the formula (3) ;
  • Step S303 calculating a superimposed volume of each pair of Gaussian spheres, wherein the ijth Gaussian sphere group includes a first Gaussian sphere and a jth Gaussian sphere, and the overlapping volume of the ijth Gaussian sphere group is ⁇ , because the Gaussian function
  • the product of the product is still a Gaussian function, so the superposition integral can be obtained simply, and the obtained superimposed integral is retained in the two-dimensional array for reuse in later calculations;
  • Step S304 calculating the weight of each Gaussian sphere, and retaining it in a one-dimensional array for later use, wherein the weight of the i-th Gaussian sphere - "" ⁇ , is the volume of the first Gaussian sphere, k is a Often
  • Step S)5 calculating the self-lamination volume of the first molecule as ⁇ n m , using the self-superimposed volume of the first molecule as the volume of the first molecule, where ⁇ is the weight of the i-th commerce sphere, ⁇ is The weight of the jth Gaussian sphere, t is the overlapping volume of the i-th Gaussian sphere and the j-th Gaussian sphere, which is the first molecular
  • Step S401 the system breaks in the first molecule and the second molecule, and calculates the volume of each molecule in the first embodiment, and the weight of each atom is retained in the computer memory.
  • the specific method is: reading the dimensional structure information of the first molecule, wherein the three-dimensional structure information includes each of the first molecules in the first molecule according to the type of each atom in the first molecule to obtain a corresponding van der Waals radius, and transforming the three-dimensional structure information
  • each Gaussian sphere the first in the molecule: the weight of the i Gaussian spheres is ⁇ 3 ⁇ 4, ⁇ ; is the i in the first molecule
  • the volume of a Gaussian sphere, k is a constant; the self-superimposed volume of the first molecule is ⁇ ⁇ > represents the set of all Gaussian spheres of the first molecule;
  • Step S402 the initial merging of the two molecules is initiated as the initial stacking of the two molecules, and the center of gravity of the two molecules is superimposed in advance, and the overlapping volume of the pair of molecules is maximized, so The initial stacking does not have to be very precise.
  • One of the simplest overlaps of the center of gravity is to translate the average position of each atom of each molecule as the center of gravity, so that they are stacked together;
  • one of the molecules (such as the first molecule) is fixed, and the other molecule is kneaded into the rigid body rotation translation transformation, so that the intermolecular superimposed volume is maximized by the calculation of the formula (8).
  • the rigid body rotation translation transformation involves six degrees of twist, in this embodiment, using Newton-Raphson
  • Step S403 calculating the intermolecular overlapping volume XW of the first molecule and the ::::: molecule for the superposition of ⁇ , wherein ⁇ ' is the ith atom in the 'th molecule and the second in the second molecule"
  • the overlapping volume of atoms calculating the first and second derivatives of the coordinates of each atom in the second molecule, and then converting the first and second derivatives of each atomic coordinate into a second molecule for overall rotational translation
  • the first and second derivatives of the variable obtain the first and second derivatives of the six degrees of freedom of the second molecule;
  • Step S404 using the Newton Rafson method to solve the change of the six degrees of freedom of the second molecule rotational translation, and coordinate transformation of the second molecule: step S405;
  • Step S405 the convergence is judged. If the convergence condition is reached, step S406 is performed, otherwise, the process returns to step S403, and the execution is continued;
  • a common problem in optimization is multiple poles, and here is the global maximum stack volume.
  • the solution to this problem is to use multiple initial folds (initial orientation) and choose the largest from the optimized results.
  • a plurality of different initial stacks may be obtained by rotating any one of the stacks along different axes.
  • Step S406 saving the intermolecular superimposed volume under the current initial superposition, and if there is a starting superposition performed by Zhu, selecting an initial superimposed ⁇ which is also performed to be the pre-initial superimposition, and executing Step S403, otherwise the maximum value in the intermolecular overlapping volume obtained under all the initial superpositions is taken as the maximum overlapping volume, and step S407 is performed; Step S407, the similarity between the first molecule and the second molecule is calculated. Similar output
  • the convergence condition of the above step S405 is usually that the modulus of the first derivative is smaller than a certain threshold.
  • the contribution of the pharmacophore to the molecular similarity comparison is increased.
  • the process is as shown in FIG. 5, and the system reads the three-dimensional structural information of the first molecule and the second molecule, and calculates the body of each molecule.
  • Each pharmacophore is represented by a Gaussian sphere with a radius of 2 angstroms and is marked with a different color depending on its type (for example, the bond donor is green, the gas bond acceptor is pink, the hydrophobic group is light blue, and the positive charge is It is red, the negative charge is dark blue, etc.).
  • the superposition of two molecules needs to be optimized.
  • the difference from the example .:::::: is that the objective function optimized here is not only the superimposed volume of two molecules but also the superimposed reward between the two drugs in the same molecule.
  • similar pharmacophores are relatively dispersed in the same molecule, so it is not necessary to consider the weight of the school: ! ⁇ .
  • Target for optimization The function is:
  • is the intermolecular superimposed volume, which is obtained by the formula (8), which is the overlapping volume between the i-th pharmacophore in the first molecule and the j-th pharmacophore in the second molecule, and The summation is limited to the same type (same color) between the pharmacophores.
  • Step S502 the two molecules are initially superposed, as the starting point of the maximum volume superposition of the two molecules, and the two molecules are superposed in advance. Translating the average position of each atom of each molecule as the center of gravity, causing them to overlap;
  • one of the molecules (such as the numerator) is determined, and the other molecule B is subjected to the rigid-body rotational translation transformation, so that the intermolecular integrated superimposed volume is calculated according to the formula (9). maximum.
  • the forward rotation translation transformation involves six degrees of freedom. In this embodiment, Newton Rafson is used.
  • the C ew on-Raphs n) method optimizes these six degrees, and performs step ⁇ 03;
  • Step S503 calculating the intermolecular integrated superimposed volume /V of the 'molecular and the second molecule for the current superposition, and calculating the first and second order of the coordinates of each atom in the second molecule
  • Step S505 the judgment is converged, if the convergence condition is reached, step S506 is performed, otherwise step S503 is performed ;
  • a common problem in optimization is multiple extremum, and what is sought here is the global maximum superimposed volume.
  • the solution to this problem is to use multiple initial folds and choose the largest from the optimized results.
  • a plurality of different initial stacks can be obtained by any - superposition of the rotation along different axes.
  • Step S506 the intermolecular integrated superimposed volume under the initial superposition is saved, and if there is still an unfinished initial superposition, then the selected initial stacking cooperation is selected. Start superimposing and perform steps
  • Step S507 calculating the similarity between the first and second molecules I '" , formula (10),
  • the convergence condition of the above steps is usually that the modulus of the first derivative is less than a certain threshold.
  • Example 4 Stacking of a group of molecules
  • Step S601 the system reads in a group of molecules, calculates the volume of each molecule, and each atom in the molecule is stepped S6()2, and then selecting one of the groups of molecules as the target target molecule may be the first The molecules, or the largest molecules, are also specified as needed. After the target molecule is selected, the molecule is fixed;
  • the set of superimposed molecules tt] thus obtained is analyzed by the ' ⁇ '.: quantitative s t : ru s t re-ac t ty relationship (QSAR) analysis or pharmacophore model.
  • QSAR quantitative s t : ru s t re-ac t ty relationship

Landscapes

  • Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Theoretical Computer Science (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computing Systems (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Medical Informatics (AREA)
  • Evolutionary Biology (AREA)
  • Biotechnology (AREA)
  • Medicinal Chemistry (AREA)
  • Pharmacology & Pharmacy (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Description

背景技术
两个分子的三维形状和药效面 (分子中具有特定性质的原子集固) 的彼此 匹配和定量比较是药物设计中主要方法 分子体积与形状相关, 面形状决定分 子的物理化学性质, 而这种性质决定分子的生物活性。 个原子, 而球的半径为原子的范德华半径。 每个硬球对应的体积密度函数在球 内为 , 而球外则为 0。 这样的体积密度可由单个原子硬球体积密度展开为下面 的公式 (1):
' ) ∑ ;0')—∑. W'/〕W+∑ fji ) ∑ 脚應' (^..
Ππ 綱
其中 fi(r)为原子 i的体积密度。如图】.所示,一个双原子分子的体积等于 两个原子体积的和减去它们之间的叠合体积,即 V^VA+VB VAB。对于.三原子分子, 则由多个 "包含" 与 "排除"交替项组成; V VA+VB+VC VAB VAOVBOV\BC。 对 于任意分子, 其体积也为一系列 "包含' f 与 "排除"交替项组成, 用下述公式 (2) 计算: V ^ j F(r) r
其中 v .为原子〖的体积, v〖 j为原子 ί和 j的 (二阶) 叠合体积, v i jk为 原子 i> j及 k的 (.Ξ阶) 叠合体积, 依此类推。 对于一个 N个原子的分子, 以 上求和一般要进行到 N 1阶原子的叠合。 运用上述方法计算分子形状的主要缺陷是: 由于硬球模型得到的体积对原 子坐标导数的不连续性, 导致计獰吋间加长, 精度下降 发明内容
本发明提供 种分子体积计算方法及两分子的形状比较方法, 以解决现有 技术计算分子体积时间较长, 精度不高的技术问题。
一种分子体积计獰方法, 所述方法包括:
歩骤 11, 读入第一分子的三维结构信息, 所述三维结构信息包括第一分子
歩骤 12, 根据第 ··· 分子中各原子的类型得到相应的范德华半径, 将三维结 构信息转换为一组代表第一分子中各原子的高斯球, 每个高斯球的半径与相应
步骤 13, 计算每对高斯球组的叠合体积, 其中第 i j 组高斯球组包括第 : i 个高斯球和第 j个高斯球, 第 i j组高斯球组的叠合体积为 ; 歩骤〗 4, 计算每个高斯球的权重, 第 i个高斯球的权重 — ~~ —, 为 第 ί个高斯球的体积, k为一个常数;
歩骤 15,计獰第一分子的自身叠合体积为 ∑ H.- w;.v ?以第一分子的自身 叠合体积作为第一分子的体积, 其中 为第 i个高斯球的权重,, 为第 j个高 斯球的权重, 为第 i.个高斯球与第 j个高.斯球的叠合体积, A为第一分子所有
迸一歩的, k的取值为 (). 5至 1 , 0之间 一种两个分子形状的比较方法, 所述方法包括: 歩骤 31, 系统读入第一分子和第二分子的: Ξ维结构信息, 计算第一分子和 第二分子的自身叠合体积, 具体包括: 读入第一分子的 Ξ维结构信息, 所述三维结构信息包括第一分子中每个原
根据第一分子中各原子的类型得到相应的范德华半径, 将≡维结构信息转 换为一组代表第一分子中各原子的高斯球, 每个高斯球的半径与相应原子的范 德华半径相同, 且每个高斯球的位 S与相应原子的坐标相同;
分子中第 i个高斯球和第 j个高斯球, W- 分子中的第 ij组高斯球组的叠合体 积 计算第一分子 每个高斯球的权靈, 第一分子中第 ; 个高斯球的权重
V.
为第一分子中第 i个高;
( v..
-;:: i.
计算第一分子的自身叠合体积为 ∑■ A表示第一分子所有高斯球
读入第:分子的三维结构信息, 所述≡维结构信息包括第二分子中每个原 子的类型及其坐标数值; 根据第二分子中各原子的类型得到相应的范德华半径, 将≡维结构信息转 换为一组代表第:二分子中各原子的高斯球, 每个高斯球的半径与相应原子的范 德华半径相同, 且每个高斯球的位置 相应原子的坐标相同; 计算第二分子中每对离斯球组的叠合体积, 其中第 ij组离斯球组包括第二 分子中第 ί个高斯球和第 j个高斯球, 第.:::::分子中的第 ϋ组高斯球组的叠合体 积为 ^; 计算第二分子中每个高斯球的权重, 第:二分子中第 ί 个高斯球的权重
Μ; ' 一 一, V为第二分子中第 ί个离斯球的体积, * '为一个常数; 计算第二分子的 ί¾身叠合体积为 Τ «ν,,' , Β 表示第二分子所有 ¾斯 球的集合;
歩骤 32, 计算第一分子.和第二分子在多种叠合愦况下的分子间体积, 选择 其中的最大值作为最大分子间体积;
歩骤 33, 根据第 -一分子的体积、 第二分子的体积和最大分子间体积计算第 一分子和第二分子的相似度, 以相似度怍为两个分子形状的比较结果。 进一歩的:
所述歩骤 32的分子间体积为分子间叠合体积, 具体包栝: 计算第一分子和第二分子在多种叠合情况下的分子间叠合体积 ∑ 作为分子间体积, 其中 1 "为第一分子中第 i个原子与第二分子 中第 j个原子的叠合体积, 选择其中的最大值 作为最大分子间体积; 所述歩骤 33具体包括: 计算第一分子和第二分子的相似度 : w > 以相似度 作为两个 进… 步的:
所述歩骤 32的分子间体积为第一分子和第二分子的分子闻综合叠合体积, 具体包括:
歩骤 5.1, 找出每个分子中的药效团及其位置:
歩骤 52, 确定每个分子中的每个药效团的类型;
歩骤 53, 计獰第一分子的自身综合叠合体积 ∑ / , 其中 为第一 分子中的第 i个药效团与第 j个药效团之间的叠合体积, 且对 的求和仅限于 同类药效团之间;
歩骤 ¾, 计算第二分子的 身综合叠合体积 α ί^ X ', 其中 '为第 二分子中的第 ί个药效团与第 j个药效团之间的叠合体积, 且对 '的求和仅限 于同类药效团之间;
歩骤 ,计算第一分子和第二分子的分子间综合叠合体积¾ ÷ * > 其中 为第一分子中的第 ί个药效团 第二个分子中的第 j个药效園之间的叠 合体积, 且对 Fti "的求和仅限于同类药效团之间; 歩骤 56, 计算第一分子和第二分子在多种叠合情况下的分子间綜合叠合体 积 作为分子间体积, 选择其中的最大值 012^作为最大分子间体积; 所述歩骤 33具体包括: 计算第一分子和第二分子的相似度 - . Gi Λ , 以相似度 .作为两个 分子形状的比较结果。
更迸一歩的, 所述 k的取值为 0. 5至: L 0之间 , 所述 的取值为 0. 5至 1. 0 之-间。 本发明降低分子的三维形状定量比较的计算复杂度、 提高计獰精度 本发 明可应用于分子形状的数学表达、 药物分子间形状比较、 药物分子药效國比较, 这些比较方法用予药物分子的虛拟筛选
附图说明
圏 i为分子叠合体积计算的示意图。
图 2A 是本发明计算得到的分子体积(( V)与硬球模型所得值(HSV)的对 比 圏 3是本发明计算分子体积流程圈。
图 4是本发明计算分子最大叠合及比较形状相似度的流程圈„
图 δ是本发明比较两个分子形状及药效团综合相似度的流程国。
图 6是本发明叠合一组分子的过程图。
具体实施方式 本发明提出的一种分子的三维形状定量比较方法, 它降低分子的三维形状 定量比较的计算复杂度、 提高计算精度
对于公式 ( 1. ) 和公式 (2 ) 采用高斯函数 取代硬球密度函数 f&d, 而公式 (i ) 和公式 (2 ) 的其余部分保持不变。 对应于分子中每个原子的高斯 体积密度 : g r}
Figure imgf000008_0001
(3)。
其中 r为任意一点坐标, .rf为原子 的坐标, σ '为该原子的范德华半径, 而 β为一可调參数, 通常取值为 2. 7。 由于高斯函数易于求积分和导数, 降低了计 算复杂度 但是, 运用公式 (2 ) 计算分子形状时, 高阶项的求和数目呈组合爆 i,因此, 对公式 (:υ作如下近似, 即将分子 体积密度简化为原子体积密垵
G(r) 公式 (4) 这个近似简化了计算, 但由于完全忽略了原子之间的体积叠合, 降-
Figure imgf000009_0001
公忒(4)计算分子自身
Figure imgf000009_0002
2—6 之间。 由于计算分子形状相似度涉及分子间和分子自身的叠合
5. !2 , 公式 (5), 采用公式 (4)给相似度计算带来大的误莲 通过对公式 (4)的进一歩修正: 得到一种分子的三维形状定量比较方法。
对于住意一个由 个原子组成的分子, 其形状 ]'用一个体积密度函数表达: 通过下式计獰:
Figure imgf000009_0003
4(ί!' ' , 公式 (6λ 其中 为原子 ./的权:盧因子, .r为任意一点坐标, 为歸 ί子
该原子的范德华半径, 而可调参数 在这凰取值为 2^, 这样的取值
的叠合体积等于原子本身体积, 即 4?κ^ /3。
键是
Figure imgf000009_0004
计算出的分子 il身叠合体积 ι 于其正确值数倍。 权重因子是原子在分子 V-U 境有关的 ·····个经验函数, 这个经验函数满足如下条件: 其权重应为 h 如果与其它原子有叠合, 其值小于 ί., 且叠合越大, 则权值越小 本发明提出如下方法确定权值 - 一 ~ —, 公式 (7) 其中 为一普适经验参数, 可通过拟合得到: k
' v t k v .. 在 0. 5至: L 0之间 依此, 两个分子的分子间叠合体积按下式计算:
∑ wswf 公式(8)。其中, 为第一分子的第 i个原子的权重, w; ' 为第二分子的第 j个原子的权重, 1 "为第一分子的第 i个原子与第二分子的第 j个原子的叠合体积, A表示第一分子所有原子的集合, B表示第:二分子所有原 子的集合。
如果第一分子和第二分子的三维结构相同, 那么 ^变成了分子自身的叠合 体积, 它应该与分子本身的体积相等 以上公式中的 k常数就是通过一组分子 自身叠合体积与分子通过硬球模型得到的体积拟合而得。 图 2 A显示了通过公式 (8)得到的分子 ί身叠合体积与分子的实际体积的对比关系, 两者的相关性趋 近于 i。 图 2B显示了文献中简单加和方法得到的自身叠合体积 分子实际体积
相关性远低于阁 2A中本发明所得到的相关性。
由于本发明提高了分子间叠合体积的精度, 因此由公式 (5 ) 计算分子间的 形状相似度的精度也获得提高。
另外, 权值也可作为 fe验参数按原子在分子中的柘扑厲性 (原子杂化状态, 成键类型和数量, 相邻原子类型等) 加以确定 这样确定的权值与分子的构象 没有关系, 这样的特性 ι¾·减少分子间的柔性叠合及分子形状相似性的柔性比较 的复杂度 实施例 ·: 分子体积计獰
如图 所示。
步骤 S301 , 读入第一分子的三维结构信息, 所述≡维结构信息包括第一分 歩骤 S302 , 根据第一分子中各原子的类型得到相应的范德华半径, 将三维 结构信息转换为一组代表第 - 4各原子的高斯球, 每个高斯球的半径与相 应原子的范德华半径相同, 且每个高斯球的位置与相应原子的坐标相同, 高斯 函数的具体表达式由公式 (3 ) 得到;
歩骤 S303, 计算每对高斯球组的叠合体积, 其中第 i j组高斯球组包括第 个高斯球和第 j个高斯球, 第 i j组高斯球组的叠合体积为^ ,因为高斯函数的 乘积仍为高斯函数, 因此叠合积分可以简单地得到, 将得到的叠合积分保留在 二维数组中 , 以便在后面的计算中重复使用;
歩骤 S304, 计算每个高斯球的权重, 并保留在一个一维数组中, 以备后用, 其中第 i个高斯球的权重 —"" ~, 为第 个高斯球的体积, k为一个常
歩骤 S )5,计算第一分子的自身叠合体积为 ∑ nm ,以第一分子的自 身叠合体积作为第一分子的体积, 其中 ^为第 i 个商斯球的权重, ^为第 j个 高斯球的权重, t 为第 i个高斯球与第 j个高斯球的叠合体积, 为第一分子所
¾ i j斯球 集合 c 实施例
Figure imgf000011_0001
歩骤 S401., 系统渎入第 -分子和第.二分子, 采用实施例一的方式计算每个 分子的 ϋ身叠合体积, 每个原子的权值, 并保留于计算机内存中。 具体方式为: 读入第 -分子的≡维结构信息, 所述三维结构信息包括第一分子中每个原 根据第一分子中各原子的类型得到相应的范德华半径, 将三维结构信息转
计算第一分子中每对高斯球组的叠合体积, 其中第 i j组¾斯球组包括第一 分子中第 i个高斯球和第 j个高斯球, 第 分子中的第 ί j组高斯球组的叠合体 计算第一分子 E:!:s每个高斯球的权重, 第 ····'分子中第 : i 个高斯球的权重 一~ ¾一, ν;为第一分子中第 i个高斯球的体积, k为一个常数; 计獰第一分子的自身叠合体积为 ∑■ Μ^ > 表示第 -分子所有高斯球 的集合;
读入第:二分子的三维结构信息, 所述三维结构信息包括第二分子中每个原 裉据第二分子中各原子的类型得到相应的范德华半径, 将三维结构信息转 换为一组代表第二分子中各原子的高斯球, 每个高斯球的半径与相应原子的范 德华半径相同, 且每个高斯球的位置与相应原子的坐标相同; 分子中第 i个高斯球和第 j个高斯球, 第二分子中的第 ί j组高斯球组的叠合体 计獰第.二分子中每个高.斯球的权重, 第.二分子中第 〖 个高新球的权重 Μ,' -一"" V,为第二分子中第 i个高斯球的体积, 为一个常数; 计算第二分子的自身叠合体积为 ∑ ',B 表示第二分子所有高斯 球的集合;
歩骤 S402, 两个分子迸行起始叠合, 作为两个分子最大体积叠合的起始, 事先将两个分子的重心叠合 由子对分子的叠合体积要进行最大化的优化, 因 此起始叠合不必十分精确。 一个最简单的重心叠合就是将每个分子各原子的平 均位置作为重心进行平移, 使它们叠合在一起;
在叠合优化时将其中的一个分子 (如第一分子) 固定, 将另一个分子 Β进 行刚体旋转平移变换, 使分子间叠合体积按公式 (8) 的计算达到最大。 刚体旋 转平移变换涉及六个 ώ度, 在本实施例中, 采用牛顿 -拉夫森
( ewton-Raphson ) 方法对这六个¾由度进行优化, 执行歩骤 S403;
歩骤 S403, 对于 ^前的叠合, 计算第一分子和第:::::分子的分子间叠合体积 X W, 其中 ν '为第 '分子中第 i个原子与第二分子中第」个原子的 叠合体积, 计算^对第二分子中的每个原子的坐标的一阶和二阶导数, 然后将 各原子坐标的一阶.和二阶导数转换成第二分子进行整体旋转平移变量的一阶和 二阶导数, 得到^对第二分子旋转平移六个自由度的一阶和二阶导数;
歩骤 S404,利用牛顿 拉夫森方法求解变换第二分子旋转平移六个自由度的 变化值, 并依此对第二分子进行坐标变换:, 执行步骤 S405;
歩骤 S405, 判断收敛, 如果达到收敛条件, 执行歩骤 S406, 否则回到歩骤 S403 , 继续执行;
优化中一个常见的问题是多重极値, 而这里要寻求的是全局最大叠合体积 解决这一问题的方案是釆用多个起始叠合(初始取向), 并从优化后的结果中选 择最大的。 通常, 多个不同的起始叠合可 ώ任意一个叠合经过沿不同的轴旋转 而得。
歩骤 S406, 保存当前起始叠合下的分子间叠合体积, 如果还有朱执行的起 始叠合, 则选择一个还来执行的起始叠合怍为 前起始叠合, 并执行歩骤 S403, 否则在所有的起始叠合下得到的分子间叠合体积中的最大值 作为最大叠合 体积, 执行歩骤 S407; 歩骤 S407, 计算第一分子和第二分子的相似度 输出相似
Figure imgf000014_0001
度 作为两个分子形状的比较结果。
上述歩骤 S405的收敛条件通常为一阶导数的模小于某一设定的阈值。
实施例三增加了药效团对分子相似度比较的贡献, 其过程如 ϋ 5所示, 包 歩骤 S501 , 系统读入第一分子和第二分子的三维结构信息, 计算每个分子 的 身叠合体积, 每个原子的权值, 找出每个分子中的药效团及其位置, 并将 这些结果保留于计算机内存中。 每个药效团用一个半径为 2 埃的高斯球表示, 并根据其类型标以不同颜色 (例如, 键给体为绿色, 氣键受体为粉红色, 疏 水基团为浅蓝, 正电荷为红色, 负电荷为深蓝等)。 为获得综合相似度, 两个分 子的叠合需要优化。 与例.::::::所不同的是, 这里进行优化的目标函数除了两分子 的叠合体积外还要加上两个分子中同类药效 I 之间的叠合赏献 由于 -般而言 同类药效团在同一分子中相对分散, H此不必考虑权重校 :!Ε。 用于优化的目标 函数为:
o ^vl2 + ιγ' , 公式( 9 )。其中 ^为分子间叠合体积, 由公式( 8 )得到, 为第一分子中的第 i个药效团与第二个分子中的第 j个药效团之间的叠合体 积, 且对 的求和仅限于同类 (同色) 药效团之间。
由公式(9 )计算得到第一分子的 身综合叠合体积 A + ∑ Fv , 其中 . 为第 分子中的第 ί个药效团与第 j个药效团之间的叠合体积, 且对^的求和 仅限于同类药效园之间, 计算得到第二分子的自身综合叠合体积 ο, - -ί- ∑ iy , 其中. 为第二分子中的 II i 个药效团与 II j个药效画:!之间的 叠合体积, 且对! '的求.和仅限 : 同类药效闭之间;
歩骤 S502, 两个分子进行起始叠合, 作为两个分子最大体积叠合的起始, 事先将两个分子的靈心叠合。 将每个分子各原子的平均位置作为重心进行平移, 使它们叠合在 - 起;
在叠合优化时将其中的一个分子 (如第 ······分子) 闘定, 将另一个分子 B进 行刚体旋转平移变换, 使分子间综合叠合体积按公式 (9 ) 的计算达到最大。 刚 傳旋转平移变换涉及六个自由度, 在本实施例中, 釆用牛顿 拉夫森
C ew on-Raphs n) 方法对这六个 由度迸行优化, 执行步骤 ^03;
歩骤 S503, 对于.当前的叠合, 计算第 '分子和第二分子的分子间综合叠合 体积 /V, 计算 对第二分子中的每个原子的坐标的一阶和二阶导
Figure imgf000015_0001
数, 然后将各原子坐标的一阶和二阶导数转换成第二分子进行整体旋转平移变 量的一阶和二阶导数, 得到 对第二分子旋转平移六个自由度的一阶和二阶导
歩骤 利用牛顿 拉夫森方法求解变换第二分子旋转平移六个 si ώ度的 变化值, 并依此对第二分子迸行坐标变换, 执行歩骤 S5()5;
歩骤 S505, 判断收敛, 如果达到收敛条件, 执行步骤 S506, 否则执行步骤 S503;
优化中一个常见的问题是多重极值, 而这里要寻求的是全局最大叠合体积。 解决这一问题的方案是釆用多个起始叠合, 并从优化后的结果中选择最大的。 通常, 多个不同的起始叠合可由任意 - 叠合经过沿不同的轴旋转而得。
歩骤 S506, 保存 前起始叠合下的分子间综合叠合体积, 如果还有未执行 的起始叠合, 则选择······个还未执行的起始叠合作为当前起始叠合, 并执行步骤
S503, 否则在所有的起始叠合下得到的分子间綜合叠合体积中的最大值 ^作 为最大叠合体积, 执行歩骤
歩骤 S507,计算第 ······分子和第二分子的相似度 I '" , 公式(10),
( ; ( . 输出相似度 作为两个分子形状的比较结果。
上述步骤 的收敛条件通常为一阶导数的模小于某一设定的阈值。 实施例四: 一组分子的叠合
本例是实施例二的扩充, 流程见图 6, 包括:
歩骤 S601, 系统读入一组分子, 计算每个分子的体积, 每个分子中各原子 歩骤 S6()2, 然后从这组分子中选择 -个作为目标 目标分子的选取可以是 第一个分子, 或体积最大的分子, 也是根据需要特别指定。 选定好目标分子后, 让该分子固定;
歩骤 目标分子以外的其余分子通过实施例=公开的方式迸行旋转平 移与 标分子达到最大叠合。 由于多极值问题, 对于每个叠合分子可采用多个 起始叠合位置, 然后选取最佳的叠合结果;
歩骤 将 标分子和. 之叠合好的分子坐标输出。 这样得到的一组叠 合分子 tt]用 '·〕··' .:Ξ維定量拘效'方法 ( quantitative s t :ruc t u re-ac t i ty relationship, QSAR) 分析或者药效团模型的建立。

Claims

权 利 要 求 书
1. 一种分子体积计獰方法, 其特征在于, 所述方法包括 - 步骤(:1.1 ), 读入第一分子的三维结构信息, 所述≡维结构信息包括第一分
歩骤 (i2), 根据第一分子中各原子的类型得到相应的范德华半径, 将三维 结构信息转换为一组代表第 - ί:μ各原子的高斯球, 每个高斯球的半径与相 应原子的范德华半径相同, J.每个高斯球的位置与相应原子的坐标相同;
个¾斯球和第 j个 斯球, 第 ί j组高斯球组的叠合体积为 ; 歩骤(1.4) , 计算每个高斯球的权重, 第 i个高斯球的权重^
Figure imgf000018_0001
为第 i个高斯球的体积, k为一个常数;
歩骤 (15 ), 计算第一分子的自身叠合体积为 ∑ w„,以第一分子的 自身叠合体积作为第一分子的体积, 其中 ^为第 : i 个高斯球的权靈, ^为第 j 个高斯球的权重, v.,为第 ί个高斯球与第 j个高斯球的叠合体积, 为第一分子 所有离斯球的集合。
2. 根据权利要求 1所述的分子体积计算方法,其特征在于, k的取值为 0. 5 至 1. 0之间。
3. 一种两个分子形状的比较方法 f 其特征在于, 所述方法包括:
歩骤 (31 ), 系统读入第一分子和第二分子的三维结构信息, 计算第一分子 和第二分子的自身叠合体积, 具体包括;
渎入第一分子的三维结构信息, 所述三维结构信息包括第 ··· 分子中每个原 子的类型及其坐标数值; 一分子中各原子的高斯球, 每个离斯球的牟径与相应原子的范 德华半径相同, 且每个高斯球的位置与相应原子的坐标相同;
计算第一分子中每对高斯球组的叠合体积, 其中第 ij组高斯球组包括第一 分子中第 i个高斯球和第 j个高斯球, 第一分于中的第 ί j组商斯球组的叠
Figure imgf000019_0001
的集合;
Figure imgf000019_0002
Figure imgf000019_0003
计算第二分子中每对高斯球组的叠合体积, 其中第 ij组高斯球组包括第二 分子中第 i个高斯球和第 j个高斯球, 第::::分子中的第 ij组高斯球组的叠合体 计算第二分子中每个高斯球的权重, 第二分子中第 i 个;
:分子中第 i个高斯球的体积, A'为一个常数;
Figure imgf000019_0004
•Pt舅. ■ -- ^.: :积为 X w^w 'v^ , Β 表示第二分子 球的集合;
歩骤(.32), 计算第一分子和第二分子在多种叠合情况下的分子间体积, 选 择其中的最大值作为最大分子间体积;
歩骤 (;¾), 根据第一分子的体积、 第二分子的体积和最大分子间体积计算 第一分子和第二分子的相似度, 以相似度作为沔个分子形状的比较结果。
4. 根据权利要求 3所述的一种两个分子形状的比较方法, 其特征在 :f: 所述歩骤 (: ) 的分子间体积为分子间叠合体积, 具体包括:
计算第 ······分子和第::分子在多种叠合情况 F的分子间叠合体积
Vi2 ^ ∑ ^' 作为分子闻体积, 其中 .为第一 ·分子中第 i个原子与第二分子 中第 j个原子的叠合体积, 选择其中的最大值^ ^作为最大分子间体积;
所述歩骤 (33) 具体包括: 计算第一分子和第:::分子的相似度 -—"- ,„ , 以相似度 ^作为两个
5. 根据权利要求 3所述的一种两个分子形状的比较方法, 其特征在 所述步骤 ( 32 ) 的分子间体积为第 分子和第二分子的分子间综合叠合体 积, 具体包括:
歩骤 (51), 找出每个分子中的药效团及其位置;
歩骤 (52), 确定每个分子中的每个药效团的类型;
歩骤(53), 计算第一分子的 身综合叠合体积 Ο, ·;+ ∑ F , 其中 为第 一分子中的第 i个药效团与第 j个药效团之间的叠合体积, 且对 的求和仅限 • 同类药效团之间;
歩骤 4), 计算第二分子的自身綜合叠合体积 ( ,其中 '为
Figure imgf000020_0001
第二分子中的第 i个药效团与第 j个药效团之间的叠合体积, 且对 '的求和仅 限于同类药效团之间;
步骤 ( 55 ), if 算:第一分子和第 分子的分子间综合叠合体积 Olt - Va i- ∑ f , 其中 为第一 -分子中的第 i个药效 ffl.与第.二个分子中的第 j 个药效团之间的叠合体积, 且对 "的求和仅限于同类药效团之间;
歩骤(56 ), 计算第一分子和第二分子在多种叠合情况下的分子间综合叠合 体积 0 ^作为分子间体积, 选择其中的最大值 Oi2 作为最大分子间体积;
所述歩骤 (33.) 具体包括- 计獰第一分子和第二分子的相似度 ¾ , 以相似度 作为两个
Oi ( …
分子形状的比较结果
6. 裉据权利要求 ίΓ'δ住一项所述的一种两个分子形状的比较方法, 其特征 在于, 所述 k的取值为 0, 5至 1, 0之间, 所述 的取值为 0, 5至 1. ()之间。
PCT/CN2012/084713 2012-07-16 2012-11-16 一种分子体积计算方法及两分子的形状比较方法 Ceased WO2014012309A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
JP2014545073A JP2015502611A (ja) 2012-07-16 2012-11-16 分子体積の計算方法及び2つの分子の形状の比較方法
US14/351,978 US9811642B2 (en) 2012-07-16 2012-11-16 Methods for shape comparison between drug molecules
EP12881232.8A EP2752784A4 (en) 2012-07-16 2012-11-16 MOLECULAR VOLUME CALCULATION METHOD AND METHOD FOR COMPARING FORMS OF TWO MOLECULES

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201210245107.3A CN102799779B (zh) 2012-07-16 2012-07-16 一种药物分子形状的比较方法
CN201210245107.3 2012-07-16

Publications (1)

Publication Number Publication Date
WO2014012309A1 true WO2014012309A1 (zh) 2014-01-23

Family

ID=47198887

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2012/084713 Ceased WO2014012309A1 (zh) 2012-07-16 2012-11-16 一种分子体积计算方法及两分子的形状比较方法

Country Status (5)

Country Link
US (1) US9811642B2 (zh)
EP (1) EP2752784A4 (zh)
JP (1) JP2015502611A (zh)
CN (1) CN102799779B (zh)
WO (1) WO2014012309A1 (zh)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107657146B (zh) * 2017-09-20 2020-05-05 广州市爱菩新医药科技有限公司 基于三维子结构的药物分子比较方法
CN108205613A (zh) * 2017-12-11 2018-06-26 华南理工大学 一种化合物分子间相似度的计算方法及系统与它们的应用
CN108010101A (zh) * 2017-12-18 2018-05-08 重庆大学 一种绘制分子片段2d投影图的方法
CN108875298B (zh) * 2018-06-07 2019-06-07 北京计算科学研究中心 基于分子形状匹配的药物筛选方法
CN109637596B (zh) * 2018-12-18 2023-05-16 广州市爱菩新医药科技有限公司 一种药物靶点预测方法
CN109637673A (zh) * 2018-12-18 2019-04-16 广州市爱菩新医药科技有限公司 一种药物虚拟筛选方法
CN109712677A (zh) * 2018-12-21 2019-05-03 广州市爱菩新医药科技有限公司 基于数据库的三维结构检索方法、电子设备及存储介质
CN113421610B (zh) * 2021-07-01 2023-10-20 北京望石智慧科技有限公司 一种分子叠合构象确定方法、装置以及存储介质
CN113593655B (zh) * 2021-08-10 2024-03-26 北京中大唯信科技有限公司 用于药物虚拟筛选的分子三维相似度的打分方法
CN116612825B (zh) * 2023-07-19 2023-10-13 四川省产品质量监督检验检测院 分子静电势等值面点云的碰撞点检测和碰撞体积计算方法

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2003019140A2 (en) * 2001-08-23 2003-03-06 Deltagen Research Laboratories, L.L.C. Method for molecular subshape similarity matching
US20030167159A1 (en) * 2001-10-04 2003-09-04 California Institute Of Technology Method for determining molecular surface areas and volumes

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002324095A (ja) * 2001-04-26 2002-11-08 Nippon Zoki Pharmaceut Co Ltd 分子類似性評価方法及びプログラム

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2003019140A2 (en) * 2001-08-23 2003-03-06 Deltagen Research Laboratories, L.L.C. Method for molecular subshape similarity matching
US20030167159A1 (en) * 2001-10-04 2003-09-04 California Institute Of Technology Method for determining molecular surface areas and volumes

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
GUO, MING ET AL.: "Study on the calculation method for molecular volume of chemical compounds.", JOURNAL OF ZHEJIANG UNIVERSITY., vol. 30, no. 5, September 2003 (2003-09-01), pages 554 - 560, XP008174071 *
See also references of EP2752784A4 *
SHANG, ZHICAI ET AL.: "Monte carlo simulation calculation of volume and surface area of molecule.", ACTA PHYSICO-CHIMICA SINICA., vol. 13, no. 12, December 1997 (1997-12-01), pages 1097 - 1100, XP055147691 *

Also Published As

Publication number Publication date
EP2752784A1 (en) 2014-07-09
JP2015502611A (ja) 2015-01-22
CN102799779A (zh) 2012-11-28
US9811642B2 (en) 2017-11-07
CN102799779B (zh) 2015-04-22
US20140278284A1 (en) 2014-09-18
EP2752784A4 (en) 2015-06-10

Similar Documents

Publication Publication Date Title
WO2014012309A1 (zh) 一种分子体积计算方法及两分子的形状比较方法
Thomas et al. Artificial intelligence in vaccine and drug design
Baranwal et al. Struct2Graph: A graph attention network for structure based predictions of protein–protein interactions
CN111785320A (zh) 基于多层网络表示学习的药物靶标相互作用预测方法
CN107209813B (zh) 用于分子结构的输入集合的相互作用参数
Savelyev et al. Molecular renormalization group coarse-graining of polymer chains: application to double-stranded DNA
Wang et al. QSAR models on aminopyrazole-substituted resorcylate compounds as Hsp90 inhibitors
Kotelnikov et al. Sampling and refinement protocols for template-based macrocycle docking: 2018 D3R Grand Challenge 4
Bakare et al. Identification of biomarkers for the accurate and sensitive diagnosis of three bacterial pneumonia pathogens using in silico approaches
Bensberg et al. Machine learning-enhanced calculation of quantum-classical binding free energies
Xia et al. The prediction of RNA-small-molecule ligand binding affinity based on geometric deep learning
Tohidinia et al. Functional exposed amino acids of CarO analysis as a potential vaccine candidate in Acinetobacter Baumannii
WO2022146632A1 (en) Protein structure prediction
Vorobjev Blind docking method combining search of low‐resolution binding sites with ligand pose refinement by molecular dynamics‐based global optimization
Reddy et al. Insights into structure and function of 30S Ribosomal Protein S2 (30S2) in Chlamydophila pneumoniae: A potent target of pneumonia
Ezhilarasan et al. In silico identification of potential drug targets in Clostridium difficile R20291: modeling and virtual screening analysis of a candidate enzyme MurG
Jha et al. Subtractive proteome analysis of Candida albicans divulges promising antifungal targets
WO2023094335A1 (en) Designing proteins by jointly modeling sequence and structure
Alazmi et al. Immuno-informatics based peptides: an approach for vaccine development against outer membrane proteins of Pseudomonas genus
Ahmad et al. Conformational dynamics of response regulator RegX3 from Mycobacterium tuberculosis
Ferrada et al. A knowledge-based potential with an accurate description of local interactions improves discrimination between native and near-native protein conformations
Krishna et al. Homology modeling and structural analysis of DNA binding response regulator of Bacillus anthracis
Lotfi et al. Determination and validation of PilT structure and ligand binding site in Neisseria gonorrhoeae: an in-silico study for vaccine and drug-targeted therapy
Bhatti Computational analysis of PAFR-targeted drug delivery
Zimmermann Molecular Modeling is an Enabling Approach to Complement and Enhance Channelopathy Research

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12881232

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2012881232

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2014545073

Country of ref document: JP

Kind code of ref document: A

WWE Wipo information: entry into national phase

Ref document number: 14351978

Country of ref document: US

NENP Non-entry into the national phase

Ref country code: DE