KRAS protein conformation change simulation method and system based on multi-sample analysis

Abstract

The invention discloses a KRAS protein conformation change simulation method and system based on multi-sample analysis, and belongs to the technical field of multi-sample analysis. The method comprises the following steps: collecting multi-sample molecular dynamics simulation data, calculating energy difference between conformations, and constructing an energy difference mapping matrix; then, based on a joint criterion of an energy difference mean value and an energy gradient change rate, a hierarchical clustering algorithm is adopted to divide a conformation energy space into a main steady-state region, a metastable-state region and a transition-state region, and a conformation hierarchical tree structure is formed; and extracting a key channel and a stable sub-state of conformation transformation according to the relationship between the energy fluctuation characteristics and transformation paths of various intervals, thereby realizing quantitative description and energy path identification of the KRAS protein conformation change. According to the method, dynamic conformation relations under different energy levels are effectively distinguished, a new calculation and analysis means is provided for conformation regulation and control and drug binding site recognition of a complex biomolecular system, and the method has the advantages of being high in accuracy, high in universality and good in expandability.

Description

Technical Field

[0001] This invention relates to the field of multi-sample analysis technology, specifically to a method and system for simulating KRAS protein conformational changes based on multi-sample analysis. Background Technology

[0002] In recent years, with the development of structural biology and computational simulation techniques, researchers have been able to reveal the conformational changes of KRAS proteins under different physiological conditions through molecular dynamics simulations, quantum chemical calculations, and conformational sampling techniques. In particular, the local motions of the GTP-binding region, P-loop, and Switch I / II regions are considered to directly affect the active state of KRAS proteins. However, existing studies mostly rely on simulation analyses of single or small samples, using time-series trajectories or principal component analysis (PCA) to describe conformational changes, making it difficult to characterize the global features of KRAS protein conformational energy distribution across multiple sample dimensions. Furthermore, traditional methods typically only consider the overall potential energy surface or principal mode energy differences, neglecting the hierarchical contributions of local residue energy changes and hydrogen bonds, van der Waals forces, and electrostatic interactions to conformational stability, leading to uncertainties in identifying steady-state and metastable states. Therefore, there is an urgent need for a simulation analysis method that can systematically calculate local energy distributions, compare energy similarities between conformations, and determine steady-state clustering at the multi-sample level to more accurately reveal the conformational change patterns of KRAS.

[0003] While existing technologies have made some progress in molecular dynamics and energy distribution analysis to address the aforementioned issues, they still have several shortcomings. First, conformational energy curves of single samples or single trajectories cannot reflect the energy stratification characteristics among multiple samples, making conformational clustering results susceptible to sampling bias. Second, existing normalization or clustering algorithms mostly use overall energy as a feature variable, lacking fine-grained comparisons of residue-level energy changes, making it difficult to distinguish between metastable and transitional states of KRAS conformations. Third, traditional energy analysis methods often ignore the mutual coupling effects between different regions when mapping between samples, failing to form a globally consistent energy similarity matrix, resulting in unstable clustering levels and inaccurate steady-state division. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for simulating KRAS protein conformational changes based on multi-sample analysis, so as to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] A method for simulating conformational changes of KRAS protein based on multi-sample analysis is disclosed. The method includes the following steps: Step S1: Extract key conformational parameters for each sample from a multi-sample dataset of KRAS protein; calculate the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the sample, and construct a local energy hierarchy matrix; Step S2: Calculate the total energy of a single residue region in the local energy hierarchy matrix of a single sample; calculate the energy difference between two adjacent residue regions; obtain the energy difference between all residue regions within a single sample, and construct a sample energy mapping matrix; Step S3: Construct a normalized sample energy mapping matrix for a single sample, and define a full mapping matrix for any two sample pairs; obtain the normalized energy difference of all sample pairs, and calculate the overall energy similarity between samples based on the normalized energy difference of all residue regions in each sample pair; Step S4: Set a preset threshold, analyze and judge sample similarity, group highly similar samples into the same cluster, calculate the average total energy for samples within each cluster; construct a conformational hierarchy tree under multi-sample analysis based on the average energy difference between clusters and sample similarity.

[0007] As a preferred embodiment of the KRAS protein conformational change simulation method based on multi-sample analysis described in this invention, a multi-sample dataset of KRAS protein is constructed, and key conformational parameters for each sample are extracted from the dataset. These key conformational parameters include the spatial angle between the α-helix H1 and the β-chain B2, the residue spacing between the P-loop and the Switch II region, and the local volume parameter of the GTP-binding cavity. A set of key conformational parameters is constructed, denoted as... ,in, This represents the spatial angle parameter of the i-th sample. This represents the residue spacing parameter of the i-th sample. Let I represent the local volume parameter of the i-th sample, and let I represent the total number of samples.

[0008] Based on the parameters in the key conformational parameter set, the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the i-th sample are calculated, and a local energy hierarchy matrix is ​​constructed, denoted as:

[0009] ;

[0010] in, This represents the local energy stratification matrix of the i-th sample. This represents the m-th residue region in the i-th sample. electrostatic energy, This represents the m-th residue region in the i-th sample. Van der Waals can, This represents the m-th residue region in the i-th sample. The hydrogen bond energy.

[0011] As a preferred embodiment of the KRAS protein conformational change simulation method based on multi-sample analysis described in this invention, the local energy stratification matrix of the i-th sample is calculated. The m-th residue region The total energy, that is ;

[0012] The energy difference between two adjacent residue regions is calculated using the following formula: ,in, Represents the local energy stratification matrix of the i-th sample. The m-th residue region and the nth residue region The energy difference between them Represents the local energy stratification matrix of the i-th sample. The nth residue region The total energy;

[0013] Obtain the energy difference between all residue regions within the i-th sample, and construct the sample energy mapping matrix for the i-th sample, denoted as . .

[0014] As a preferred embodiment of the KRAS protein conformational change simulation method based on multi-sample analysis described in this invention, sample-by-sample minimum-maximum normalization is employed for the m-th residue region in the i-th sample. The total energy is normalized, and based on the normalized total energy, a normalized sample energy mapping matrix for the i-th sample is constructed, denoted as . ;

[0015] For any two sample pairs, define the full mapping matrix for the sample pairs as follows:

[0016] ;

[0017] in, This represents the region of the m-th residue in the i-th sample. The normalized energy difference between the j-th sample and the n-th residue region;

[0018] Obtain the normalized energy difference for all sample pairs. Based on the normalized energy difference of all residue regions in each sample pair, calculate the overall energy similarity between samples using the following formula:

[0019] ;

[0020] in, This represents the overall energy similarity between the i-th sample and the j-th sample;

[0021] Obtain the overall energy similarity among all samples and construct a sample similarity matrix.

[0022] As a preferred embodiment of the KRAS protein conformational change simulation method based on multi-sample analysis described in this invention, a preset overall energy similarity threshold is established. If the overall energy similarity between the i-th sample and the j-th sample is... If the similarity is greater than the overall energy similarity threshold, then the i-th sample and the j-th sample are determined to have high similarity.

[0023] Based on the sample similarity matrix, samples with high similarity are grouped into the same cluster, and the average total energy is calculated for samples within each cluster.

[0024] Clusters with the smallest energy fluctuations are labeled as the main steady-state region, clusters with medium energy fluctuations are labeled as the metastable region, and clusters with large energy fluctuations are labeled as the transition state region.

[0025] Based on the average energy difference between clusters and sample similarity, a conformational hierarchical tree is constructed under multiple samples. The top layer of the conformational hierarchical tree is the main steady-state region, the middle layer of the tree is the metastable region, and the leaf nodes of the tree are the transitional region.

[0026] A KRAS protein conformational change simulation system based on multi-sample analysis. The system includes: a data acquisition and matrix construction module, an energy difference calculation and mapping matrix construction module, a full mapping matrix construction and similarity calculation module, and an analysis and hierarchical tree construction module.

[0027] The data acquisition and matrix construction module extracts key conformational parameters for each sample from a multi-sample dataset of the KRAS protein; calculates the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the sample; and constructs a local energy hierarchy matrix.

[0028] The energy difference calculation and mapping matrix construction module calculates the total energy of a single residue region in the local energy stratification matrix of a single sample; calculates the energy difference between two adjacent residue regions; obtains the energy difference between all residue regions within a single sample, and constructs a sample energy mapping matrix.

[0029] The full mapping matrix construction and similarity calculation module: constructs a normalized sample energy mapping matrix for a single sample; defines a full mapping matrix for any two sample pairs; obtains the normalized energy difference of all sample pairs; and calculates the overall energy similarity between samples based on the normalized energy difference of all residue regions of each sample pair.

[0030] The analysis and hierarchical tree construction module: presets a threshold, analyzes and judges the sample similarity, groups highly similar samples into the same cluster, calculates the average total energy for samples within each cluster, and constructs a conformational hierarchical tree for multiple samples based on the average energy difference between clusters and sample similarity.

[0031] Furthermore, the data acquisition and matrix construction module includes a data acquisition unit and a matrix construction unit;

[0032] The data acquisition unit: constructs a multi-sample dataset of KRAS protein, and extracts key conformational parameters for each sample from the multi-sample dataset. The key conformational parameters include the spatial angle parameter between the α-helix H1 and the β-chain B2, the residue spacing parameter between the P-loop and the Switch II region, and the local volume parameter of the GTP binding cavity, and constructs a set of key conformational parameters.

[0033] The matrix construction unit calculates the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the i-th sample based on the parameters in the key conformation parameter set, and constructs a local energy hierarchical matrix.

[0034] Furthermore, the energy difference calculation and mapping matrix construction module includes an energy difference calculation unit and a mapping matrix construction unit;

[0035] The energy difference calculation unit: calculates the total energy of the m-th residue region in the local energy stratification matrix of the i-th sample; and calculates the energy difference between two adjacent residue regions.

[0036] The mapping matrix construction unit: obtains the energy difference between all residue regions in the i-th sample and constructs the sample energy mapping matrix of the i-th sample.

[0037] Furthermore, the full mapping matrix construction and similarity calculation module includes a full mapping matrix construction unit and a similarity calculation unit;

[0038] The full mapping matrix construction unit: adopts sample-by-sample minimum-maximum normalization to normalize the sum of energy of the m-th residue region in the i-th sample, and constructs the normalized sample energy mapping matrix of the i-th sample based on the normalized sum of energy; for any two sample pairs, a sample pair full mapping matrix is ​​defined.

[0039] The similarity calculation unit: obtains the normalized energy difference of all sample pairs, calculates the overall energy similarity between samples based on the normalized energy difference of all residue regions of each sample pair, obtains the overall energy similarity between all samples, and constructs a sample similarity matrix.

[0040] Furthermore, the analysis and hierarchical tree construction module includes an analysis unit and a hierarchical tree construction unit;

[0041] The analysis unit: presets an overall energy similarity threshold; if the overall energy similarity between the i-th sample and the j-th sample is greater than the overall energy similarity threshold, it determines that there is a high similarity between the i-th sample and the j-th sample; based on the sample similarity matrix, it groups the highly similar samples into the same cluster, and calculates the average total energy for the samples within each cluster;

[0042] The hierarchical tree construction unit: the cluster with the smallest energy fluctuation is marked as the main stable state region, the cluster with medium energy is marked as the metastable state region, and the cluster with large energy fluctuation is marked as the transition state region; based on the average energy difference between clusters and sample similarity, a conformational hierarchical tree under multiple samples is constructed, the top layer of the conformational hierarchical tree is the main stable state region, the middle layer of the tree is the metastable state region, and the leaf nodes of the tree are the transition state regions.

[0043] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: This invention provides a method and system for simulating KRAS protein conformational changes based on multi-sample analysis. By extracting key conformational parameters of the KRAS protein from a multi-sample dataset, calculating the electrostatic energy, van der Waals energy, and hydrogen bond energy of each sample's residue regions, and constructing a local energy stratification matrix, it achieves accurate characterization of the local energy distribution of each sample. This allows for the coupled expression of the geometric and energy characteristics of protein conformation, providing a high-resolution foundation for subsequent energy difference analysis. By calculating the sum of energy in each residue region and the energy differences between them, a sample energy mapping matrix is ​​constructed, achieving a quantitative description of the energy gradient relationship within the sample, thereby revealing the laws governing energy transfer and distribution between different residues. Furthermore, by performing min-max normalization on the sample energy mapping matrix and establishing a full mapping matrix for sample pairs, the overall energy similarity between samples is calculated, achieving comparability and standardization of difference measurement for cross-sample conformational energy distribution. This ensures high discrimination stability even in the presence of experimental errors or energy scaling differences. Finally, by setting an energy similarity threshold, cluster analysis was performed on the samples, grouping highly similar samples into the same cluster. Based on the average energy and energy fluctuation characteristics within each cluster, a hierarchical conformational tree was constructed for multiple samples, achieving hierarchical division of the main steady-state region, metastable region, and transition state region. Through the organic connection of the above steps and the hierarchical mapping of energy parameters, this invention realizes a complete modeling process from extracting local energy features of a single sample to hierarchical conformation of multiple samples. It can not only accurately identify the conformational characteristics and transition pathways of KRAS proteins in different energy states, but also provide interpretable quantitative evidence for subsequent drug target identification, mutant activity prediction, and conformational stability analysis, ultimately achieving the beneficial effect of improving the accuracy of conformational change simulation and its correlation with biological significance. Attached Figure Description

[0044] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0045] Figure 1 This is a schematic diagram illustrating the steps of a method for simulating KRAS protein conformational changes based on multi-sample analysis according to the present invention.

[0046] Figure 2 This is a schematic diagram of the structure of a KRAS protein conformational change simulation system based on multi-sample analysis according to the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Please see Figure 1 In this first embodiment: a method for simulating KRAS protein conformational changes based on multi-sample analysis is provided, which includes the following steps:

[0049] Step S1: Extract key conformational parameters for each sample from the multi-sample dataset of KRAS protein; calculate the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the sample, and construct a local energy hierarchy matrix.

[0050] Specifically, a multi-sample dataset of the KRAS protein is constructed, and key conformational parameters for each sample are extracted from the dataset. These key conformational parameters include the spatial angle between the α-helix H1 and the β-chain B2, the residue spacing between the P-loop and the Switch II region, and the local volume parameter of the GTP-binding cavity. A set of key conformational parameters is constructed, denoted as... ,in, This represents the spatial angle parameter of the i-th sample. This represents the residue spacing parameter of the i-th sample. Let I represent the local volume parameter of the i-th sample, and let I represent the total number of samples.

[0051] It should be noted that α-helix H1 refers to the first α-helix segment formed at the N-terminus of the KRAS protein (generally residues 2 to 16). This region is close to the GTP binding site, and its spatial conformation has an important influence on the activation state of the KRAS protein. β-chain B2 refers to the second β-sheet segment (approximately residues 35 to 40) located in the central core region of the KRAS protein. It has a stable spatial stacking relationship with α-helix H1 and is used to maintain the protein's three-dimensional backbone.

[0052] Spatial angle parameters are geometric parameters used to quantify the relative orientation changes between the α-helix H1 and the β-chain B2, and can reflect the overall folding angle changes of the KRAS protein in different conformations (activated and inactivated states).

[0053] Based on the parameters in the key conformational parameter set, the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the i-th sample are calculated, and a local energy hierarchy matrix is ​​constructed, denoted as:

[0054] ;

[0055] in, This represents the local energy stratification matrix of the i-th sample. This represents the m-th residue region in the i-th sample. electrostatic energy, This represents the m-th residue region in the i-th sample. Van der Waals can, This represents the m-th residue region in the i-th sample. The hydrogen bond energy.

[0056] It should be noted that the KRAS protein is composed of a series of amino acids, each of which is called a residue. Each residue is numbered according to its position in the amino acid sequence. Each residue has a specific position coordinate (atomic coordinate) in three-dimensional space and is connected to adjacent residues through the main chain. The residue region in this invention refers to a local structural region composed of several adjacent residues that has relatively independent conformational characteristics in space.

[0057] In this invention, this step uses explicit conformational parameters (the angle between the α-helix H1 and the β-chain B2, the residue spacing between the P-loop and Switch II, the local volume of the GTP binding cavity, etc.) as input. Based on physicochemical terms (electrostatic energy, van der Waals energy, hydrogen bond energy), energy calculations are performed at the residue region level and organized in matrix form. This couples conformational geometry with physical energy and explicitly maps high-dimensional conformational differences into comparable energy vectors / matrices. On the one hand, it improves the physical interpretability of the conformational description, giving clear energy basis for any subsequent differential analysis (such as the source of residue energy mutations). On the other hand, it provides structured and standardized data units (local energy hierarchical matrices) for subsequent cross-sample quantitative comparisons, thereby improving the sensitivity and reproducibility of the method to small conformational differences (such as local energy changes caused by drug induction or mutation), which is beneficial for patentability and feasibility in industrial applications.

[0058] Step S2: Calculate the total energy of a single residue region in the local energy stratification matrix of a single sample; calculate the energy difference between two adjacent residue regions; obtain the energy difference between all residue regions within a single sample, and construct the sample energy mapping matrix.

[0059] Specifically, calculate the local energy stratification matrix of the i-th sample. The m-th residue region The total energy, that is ;

[0060] The energy difference between two adjacent residue regions is calculated using the following formula: ,in, Represents the local energy stratification matrix of the i-th sample. The m-th residue region and the nth residue region The energy difference between them Represents the local energy stratification matrix of the i-th sample. The nth residue region The total energy;

[0061] Obtain the energy difference between all residue regions within the i-th sample, and construct the sample energy mapping matrix for the i-th sample, denoted as . .

[0062] In this invention, the three types of energy terms are aggregated at the residue level. And calculate the energy difference between any two residue regions. This transforms the "absolute value of local energy" into a "network of energy differences between residues." This mapping matrix contains both local energy intensity information and implicit energy gradient directions, facilitating the identification of energy concentration / dispersion hotspots and boundaries. It can clearly indicate the specific residue sites and directions of energy redistribution within the molecule (e.g., which residue "transfers" energy from the steady state to the transition state), providing quantitative evidence for identifying key functional sites, predicting conformational transition pathways, and localizing drug interventions. At the same time, this difference matrix reduces the interference of overall energy scaling differences between different samples, enhancing the reliability of comparisons within the same system.

[0063] Step S3: Construct the normalized sample energy mapping matrix for a single sample, and define the full mapping matrix for any two sample pairs; obtain the normalized energy difference of all sample pairs, and calculate the overall energy similarity between samples based on the normalized energy difference of all residue regions of each sample pair.

[0064] Specifically, a sample-by-sample min-max normalization is used for the m-th residue region in the i-th sample. The total energy is normalized, and based on the normalized total energy, a normalized sample energy mapping matrix for the i-th sample is constructed, denoted as . ;

[0065] For any two sample pairs, define the full mapping matrix for the sample pairs as follows:

[0066] ;

[0067] in, This represents the region of the m-th residue in the i-th sample. The normalized energy difference between the j-th sample and the n-th residue region;

[0068] Obtain the normalized energy difference for all sample pairs. Based on the normalized energy difference of all residue regions in each sample pair, calculate the overall energy similarity between samples using the following formula:

[0069] ;

[0070] in, This represents the overall energy similarity between the i-th sample and the j-th sample;

[0071] Obtain the overall energy similarity among all samples and construct a sample similarity matrix.

[0072] In this invention, sample-by-sample minimum-maximum normalization is used to eliminate biases introduced by differences in measurement, simulation scale, or overall energy baseline between different samples, allowing the comparison to focus on the energy distribution pattern rather than absolute values. A sample-pair full mapping is defined. The similarity is obtained by summing / averaging the full mappings and then mapping them using an exponential function. This approach quantifies fine-grained differences between samples and compresses the impact of outliers through exponential transformation, improving the stability of discrimination. The generated sample similarity can serve not only as input for clustering but also as a continuous measure of conformational "morphological similarity," ensuring discriminative capability even in the presence of noise or sequencing / simulation errors. Furthermore, the normalization + full mapping strategy enhances the portability of sample comparisons across batches and experimental conditions, making it highly advantageous for integrating and comparing data from multiple sources.

[0073] Step S4: Set a preset threshold, analyze and judge the sample similarity, group samples with high similarity into the same cluster, calculate the average total energy of samples within each cluster, and construct a conformational hierarchical tree under multiple samples based on the average energy difference between clusters and sample similarity.

[0074] Specifically, a preset overall energy similarity threshold is set. If the overall energy similarity between the i-th sample and the j-th sample is... If the similarity is greater than the overall energy similarity threshold, then the i-th sample and the j-th sample are determined to have high similarity.

[0075] Based on the sample similarity matrix, samples with high similarity are grouped into the same cluster, and the average total energy is calculated for samples within each cluster.

[0076] Clusters with the smallest energy fluctuations are labeled as the main steady-state region, clusters with medium energy fluctuations are labeled as the metastable region, and clusters with large energy fluctuations are labeled as the transition state region.

[0077] Based on the average energy difference between clusters and sample similarity, a conformational hierarchical tree is constructed under multiple samples. The top layer of the conformational hierarchical tree is the main steady-state region, the middle layer of the tree is the metastable region, and the leaf nodes of the tree are the transitional region.

[0078] In this invention, this step transforms continuous similarity metrics into discrete conformational clusters. Steady-state categories are determined by calculating the average total energy and energy fluctuations within each cluster. A hierarchical tree is constructed based on the average energy difference and similarity between clusters, thus mapping the static sample set into a hierarchical structure with levels (master steady state → metastable state → transition state) and energy relationships. This hierarchical result facilitates both biological interpretation (e.g., which cluster represents the functionally active conformation) and downstream processes (e.g., designing inhibitors or stabilizers for transition states). It achieves systematic hierarchical management and visualization of the KRAS conformational space, clearly identifying the "master steady state" with the smallest energy fluctuation as a target identification benchmark, and identifying clusters with large energy amplitudes and low similarity as possible transitional pathways or induced states. This provides a directly usable hierarchical model and candidate conformation set for functional studies, mutation effect prediction, and small molecule screening.

[0079] Please see Figure 2 In this second embodiment: a KRAS protein conformation change simulation system based on multi-sample analysis is provided. The system includes: a data acquisition and matrix construction module, an energy difference calculation and mapping matrix construction module, a full mapping matrix construction and similarity calculation module, and an analysis and hierarchical tree construction module.

[0080] The data acquisition and matrix construction module extracts key conformational parameters for each sample from a multi-sample dataset of the KRAS protein; calculates the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the sample; and constructs a local energy hierarchy matrix.

[0081] The energy difference calculation and mapping matrix construction module calculates the total energy of a single residue region in the local energy stratification matrix of a single sample; calculates the energy difference between two adjacent residue regions; obtains the energy difference between all residue regions within a single sample, and constructs a sample energy mapping matrix.

[0082] The full mapping matrix construction and similarity calculation module: constructs a normalized sample energy mapping matrix for a single sample; defines a full mapping matrix for any two sample pairs; obtains the normalized energy difference of all sample pairs; and calculates the overall energy similarity between samples based on the normalized energy difference of all residue regions of each sample pair.

[0083] The analysis and hierarchical tree construction module: presets a threshold, analyzes and judges the sample similarity, groups highly similar samples into the same cluster, calculates the average total energy for samples within each cluster, and constructs a conformational hierarchical tree for multiple samples based on the average energy difference between clusters and sample similarity.

[0084] Furthermore, the data acquisition and matrix construction module includes a data acquisition unit and a matrix construction unit;

[0085] The data acquisition unit: constructs a multi-sample dataset of KRAS protein, and extracts key conformational parameters for each sample from the multi-sample dataset. The key conformational parameters include the spatial angle parameter between the α-helix H1 and the β-chain B2, the residue spacing parameter between the P-loop and the Switch II region, and the local volume parameter of the GTP binding cavity, and constructs a set of key conformational parameters.

[0086] The matrix construction unit calculates the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the i-th sample based on the parameters in the key conformation parameter set, and constructs a local energy hierarchical matrix.

[0087] Furthermore, the energy difference calculation and mapping matrix construction module includes an energy difference calculation unit and a mapping matrix construction unit;

[0088] The energy difference calculation unit: calculates the total energy of the m-th residue region in the local energy stratification matrix of the i-th sample; and calculates the energy difference between two adjacent residue regions.

[0089] The mapping matrix construction unit: obtains the energy difference between all residue regions in the i-th sample and constructs the sample energy mapping matrix of the i-th sample.

[0090] Furthermore, the full mapping matrix construction and similarity calculation module includes a full mapping matrix construction unit and a similarity calculation unit;

[0091] The full mapping matrix construction unit: adopts sample-by-sample minimum-maximum normalization to normalize the sum of energy of the m-th residue region in the i-th sample, and constructs the normalized sample energy mapping matrix of the i-th sample based on the normalized sum of energy; for any two sample pairs, a sample pair full mapping matrix is ​​defined.

[0092] The similarity calculation unit: obtains the normalized energy difference of all sample pairs, calculates the overall energy similarity between samples based on the normalized energy difference of all residue regions of each sample pair, obtains the overall energy similarity between all samples, and constructs a sample similarity matrix.

[0093] Furthermore, the analysis and hierarchical tree construction module includes an analysis unit and a hierarchical tree construction unit;

[0094] The analysis unit: presets an overall energy similarity threshold; if the overall energy similarity between the i-th sample and the j-th sample is greater than the overall energy similarity threshold, it determines that there is a high similarity between the i-th sample and the j-th sample; based on the sample similarity matrix, it groups the highly similar samples into the same cluster, and calculates the average total energy for the samples within each cluster;

[0095] The hierarchical tree construction unit: the cluster with the smallest energy fluctuation is marked as the main stable state region, the cluster with medium energy is marked as the metastable state region, and the cluster with large energy fluctuation is marked as the transition state region; based on the average energy difference between clusters and sample similarity, a conformational hierarchical tree under multiple samples is constructed, the top layer of the conformational hierarchical tree is the main stable state region, the middle layer of the tree is the metastable state region, and the leaf nodes of the tree are the transition state regions.

[0096] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0097] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for simulating KRAS protein conformational changes based on multi-sample analysis, characterized in that, The method includes the following steps: Step S1: Extract key conformational parameters for each sample from the multi-sample dataset of KRAS protein; calculate the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the sample, and construct a local energy hierarchy matrix; Step S2: Calculate the total energy of a single residue region in the local energy stratification matrix of a single sample; calculate the energy difference between two adjacent residue regions; obtain the energy difference between all residue regions within a single sample, and construct the sample energy mapping matrix; Step S3: Construct the normalized sample energy mapping matrix for a single sample; define the full mapping matrix for any two sample pairs; obtain the normalized energy difference of all sample pairs; and calculate the overall energy similarity between samples based on the normalized energy difference of all residue regions of each sample pair. Step S4: Set a preset threshold, analyze and judge the sample similarity, group samples with high similarity into the same cluster, calculate the average total energy of samples within each cluster, and construct a conformational hierarchical tree under multiple samples based on the average energy difference between clusters and sample similarity.

2. The method for simulating KRAS protein conformational changes based on multi-sample analysis according to claim 1, characterized in that, The specific implementation process of step S1 includes: A multi-sample dataset of the KRAS protein was constructed, and key conformational parameters for each sample were extracted from the dataset. These key conformational parameters included the spatial angle between the α-helix H1 and the β-chain B2, the residue spacing between the P-loop and the Switch II region, and the local volume of the GTP-binding cavity. A set of key conformational parameters was constructed, denoted as […]. ,in, This represents the spatial angle parameter of the i-th sample. This represents the residue spacing parameter of the i-th sample. Let I represent the local volume parameter of the i-th sample, and let I represent the total number of samples. Based on the parameters in the key conformational parameter set, the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the i-th sample are calculated, and a local energy hierarchy matrix is ​​constructed, denoted as: ； in, This represents the local energy stratification matrix of the i-th sample. This represents the m-th residue region in the i-th sample. electrostatic energy, This represents the m-th residue region in the i-th sample. Van der Waals can, This represents the m-th residue region in the i-th sample. The hydrogen bond energy.

3. The method for simulating KRAS protein conformational changes based on multi-sample analysis according to claim 2, characterized in that, The specific implementation process of step S2 includes: Calculate the local energy stratification matrix of the i-th sample. The m-th residue region The total energy, that is ; The energy difference between two adjacent residue regions is calculated using the following formula: ,in, Represents the local energy stratification matrix of the i-th sample. The m-th residue region and the nth residue region The energy difference between them Represents the local energy stratification matrix of the i-th sample. The nth residue region The total energy; Obtain the energy difference between all residue regions within the i-th sample, and construct the sample energy mapping matrix for the i-th sample, denoted as . .

4. The method for simulating KRAS protein conformational changes based on multi-sample analysis according to claim 3, characterized in that, The specific implementation process of step S3 includes: Using sample-by-sample min-max normalization, for the m-th residue region in the i-th sample The total energy is normalized, and based on the normalized total energy, a normalized sample energy mapping matrix for the i-th sample is constructed, denoted as . ; For any two sample pairs, define the full mapping matrix for the sample pairs as follows: ； in, This represents the region of the m-th residue in the i-th sample. The normalized energy difference between the j-th sample and the n-th residue region; Obtain the normalized energy difference for all sample pairs. Based on the normalized energy difference of all residue regions in each sample pair, calculate the overall energy similarity between samples using the following formula: ； in, This represents the overall energy similarity between the i-th sample and the j-th sample; Obtain the overall energy similarity among all samples and construct a sample similarity matrix.

5. The method for simulating KRAS protein conformational changes based on multi-sample analysis according to claim 4, characterized in that, The specific implementation process of step S4 includes: A preset overall energy similarity threshold is set. If the overall energy similarity between the i-th sample and the j-th sample is... If the similarity is greater than the overall energy similarity threshold, then the i-th sample and the j-th sample are determined to have high similarity. Based on the sample similarity matrix, samples with high similarity are grouped into the same cluster, and the average total energy is calculated for samples within each cluster. Clusters with the smallest energy fluctuations are labeled as the main steady-state region, clusters with medium energy fluctuations are labeled as the metastable region, and clusters with large energy fluctuations are labeled as the transition state region. Based on the average energy difference between clusters and sample similarity, a conformational hierarchical tree is constructed under multiple samples. The top layer of the conformational hierarchical tree is the main steady-state region, the middle layer of the tree is the metastable region, and the leaf nodes of the tree are the transitional region.

6. A KRAS protein conformational change simulation system based on multi-sample analysis, executing the KRAS protein conformational change simulation method based on multi-sample analysis as described in any one of claims 1-5, characterized in that, The system includes: a data acquisition and matrix construction module, an energy difference calculation and mapping matrix construction module, a full mapping matrix construction and similarity calculation module, and an analysis and hierarchical tree construction module; The data acquisition and matrix construction module extracts key conformational parameters for each sample from a multi-sample dataset of the KRAS protein; calculates the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the sample; and constructs a local energy hierarchy matrix. The energy difference calculation and mapping matrix construction module calculates the total energy of a single residue region in the local energy stratification matrix of a single sample; calculates the energy difference between two adjacent residue regions; obtains the energy difference between all residue regions within a single sample, and constructs a sample energy mapping matrix. The full mapping matrix construction and similarity calculation module: constructs a normalized sample energy mapping matrix for a single sample; defines a full mapping matrix for any two sample pairs; obtains the normalized energy difference of all sample pairs; and calculates the overall energy similarity between samples based on the normalized energy difference of all residue regions of each sample pair. The analysis and hierarchical tree construction module: presets a threshold, analyzes and judges the sample similarity, groups highly similar samples into the same cluster, calculates the average total energy for samples within each cluster, and constructs a conformational hierarchical tree for multiple samples based on the average energy difference between clusters and sample similarity.

7. The KRAS protein conformational change simulation system based on multi-sample analysis according to claim 6, characterized in that: The data acquisition and matrix construction module includes a data acquisition unit and a matrix construction unit; The data acquisition unit: constructs a multi-sample dataset of KRAS protein, and extracts key conformational parameters for each sample from the multi-sample dataset. The key conformational parameters include the spatial angle parameter between the α-helix H1 and the β-chain B2, the residue spacing parameter between the P-loop and the Switch II region, and the local volume parameter of the GTP binding cavity, and constructs a set of key conformational parameters. The matrix construction unit calculates the electrostatic energy, van der Waals energy, and hydrogen bond energy of each residue region in the i-th sample based on the parameters in the key conformation parameter set, and constructs a local energy hierarchical matrix.

8. The KRAS protein conformational change simulation system based on multi-sample analysis according to claim 7, characterized in that: The energy difference calculation and mapping matrix construction module includes an energy difference calculation unit and a mapping matrix construction unit; The energy difference calculation unit calculates the total energy of the m-th residue region in the local energy stratification matrix of the i-th sample. Calculate the energy difference between two adjacent residue regions; The mapping matrix construction unit: obtains the energy difference between all residue regions in the i-th sample and constructs the sample energy mapping matrix of the i-th sample.

9. The KRAS protein conformational change simulation system based on multi-sample analysis according to claim 8, characterized in that: The full mapping matrix construction and similarity calculation module includes a full mapping matrix construction unit and a similarity calculation unit; The full mapping matrix construction unit: adopts sample-by-sample minimum-maximum normalization to normalize the sum of energy of the m-th residue region in the i-th sample, and constructs the normalized sample energy mapping matrix of the i-th sample based on the normalized sum of energy. For any two sample pairs, define the sample pair full mapping matrix; The similarity calculation unit: obtains the normalized energy difference of all sample pairs, and calculates the overall energy similarity between samples based on the normalized energy difference of all residue regions of each sample pair; Obtain the overall energy similarity among all samples and construct a sample similarity matrix.

10. The KRAS protein conformational change simulation system based on multi-sample analysis according to claim 9, characterized in that: The analysis and hierarchical tree construction module includes an analysis unit and a hierarchical tree construction unit; The analysis unit: presets an overall energy similarity threshold. If the overall energy similarity between the i-th sample and the j-th sample is greater than the overall energy similarity threshold, then it is determined that there is a high similarity between the i-th sample and the j-th sample. Based on the sample similarity matrix, samples with high similarity are grouped into the same cluster, and the average total energy is calculated for samples within each cluster. The hierarchical tree construction unit is as follows: the cluster with the smallest energy fluctuation is marked as the main steady-state region, the cluster with medium energy is marked as the metastable region, and the cluster with large energy fluctuation is marked as the transitional region. Based on the average energy difference between clusters and sample similarity, a conformational hierarchical tree is constructed under multiple samples. The top layer of the conformational hierarchical tree is the main steady-state region, the middle layer of the tree is the metastable region, and the leaf nodes of the tree are the transitional region.

Images