Identification of Binding Sites in Protein Structures

Abstract

In one aspect, a method is provided for identifying residues within a particular conformation of a protein molecule that can form a stable bond with a compound. The method can include identifying one or more residues from conformations of the protein molecule that exhibit a protection factor meeting one or more thresholds for inclusion in a perturbation set. The one or more residues can be excluded from the perturbation set based at least on the spatial arrangement and / or orientation of the one or more residues. One or more molecular dynamics simulations can be performed that result in an indication of the interaction between the compound and the residues within the perturbation set. The residues within the perturbation set can be identified as those that form a stable bond between the compound and the protein molecule based at least on the results of the molecular dynamics simulations.

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims the benefit of U.S. Provisional Patent Application No. 63 / 347,798, filed on June 1, 2022, entitled "IDENTIFICATION OF BINDING SITES IN A PROTEIN STRUCTURE", the disclosure of which is hereby incorporated by reference in its entirety.

[0002] Technical Field The subject matter described herein generally relates to the identification of one or more binding sites in the conformation of a protein molecule.

Background Art

[0003] Molecular binding is a process by which two or more molecules (e.g., proteins, ligands, etc.) can interact to form a stable molecular structure. For example, molecular binding can include a chemical bond between a portion of a first molecule and a portion of a second molecule. The portion of the first molecule that binds to the second molecule can include multiple binding sites. For example, in the case of a protein - ligand complex, each binding site of the protein molecule can include a plurality of binding residues, which are residues that participate in the binding to the ligand. In various real - world applications (e.g., drug discovery, etc.), it may be necessary to identify binding sites in one molecule that can bind to another molecule. For example, the identification of binding residues in a protein molecule can be utilized in the design of ligands that can effectively bind to the protein molecule.

Summary of the Invention

[0004] A manufactured product including a system, method, and computer program product is provided for identifying binding sites in protein molecules. In some exemplary embodiments, a system including at least one processor and at least one memory is provided. The at least one memory may include program code that, when executed by the at least one processor, provides a plurality of operations. The operations may include: (a) determining a first set of perturbations of a first conformation of a protein molecule, including identifying one or more residues to include in the first set of perturbations based at least on protection factors of at least a portion of the residues in the protein molecule, and further excluding from the first set of perturbations at least a first residue for which a distance from the first set of perturbations to a second residue within the first set of perturbations does not correspond to one or more dimensions of a compound; (b) performing one or more molecular dynamics simulations to generate results indicative of interactions between a compound and a first plurality of residues within the first set of perturbations; and (c) identifying, based at least on the results of the one or more molecular dynamics simulations, the first plurality of residues within the set of perturbations as those forming a stable bond between the compound and the first conformation of the protein molecule.

[0005] In another aspect, a method for identifying a binding site in a protein molecule is provided. The method comprises: (a) determining a first perturbation set of a first conformation of the protein molecule, including identifying one or more residues to include in the first perturbation set based at least on the shielding coefficients of at least a portion of the residues in the protein molecule, and further excluding from the first perturbation set at least a first residue for which the distance from the first perturbation set to a second residue within the first perturbation set does not correspond to one or more dimensions of the compound; (b) performing one or more molecular dynamics simulations to generate results indicative of interactions between the compound and the first plurality of residues within the first perturbation set; and (c) identifying, based at least on the results of the one or more molecular dynamics simulations, the first plurality of residues within the perturbation set as forming a stable bond between the compound and the first conformation of the protein molecule.

[0006] In another aspect, a computer program product is provided that includes a non-transitory computer-readable medium storing instructions that, when executed by at least one data processor, can cause operations to be performed. The operations can include: (a) determining a first perturbation set of a first conformation of the protein molecule, including identifying one or more residues to include in the first perturbation set based at least on the shielding coefficients of at least a portion of the residues in the protein molecule, and further excluding from the first perturbation set at least a first residue for which the distance from the first perturbation set to a second residue within the first perturbation set does not correspond to one or more dimensions of the compound; (b) performing one or more molecular dynamics simulations to generate results indicative of interactions between the compound and the first plurality of residues within the first perturbation set; and (c) identifying, based at least on the results of the one or more molecular dynamics simulations, the first plurality of residues within the perturbation set as forming a stable bond between the compound and the first conformation of the protein molecule.

[0007] In some variations, one or more features, including the following features disclosed herein, may optionally be included in any combination that can be realized.

[0008] In some variations, the protection factor associated with each residue in the first conformation may correspond to the difference between the first energy of the residue in the bound state and the second energy of the residue in the unbound state.

[0009] In some variations, one or more residues are identified as being included in the first perturbation set based on the protection factors of one or more residues that satisfy at least one or more thresholds.

[0010] In some variations, the protection factor of one or more residues may be determined to satisfy one or more thresholds based at least on the value of the protection factor being within a percentile of the maximum protection factor value observed in the first conformation of the protein molecule.

[0011] In some variations, the percentile may be determined based at least on the magnitude of the difference between the maximum protection factor value and a plurality of other protection factor values observed in the first conformation of the protein molecule.

[0012] In some variations, the percentile may be a value between 5 and 20.

[0013] In some variations, the interaction between the compound and the first plurality of residues in the first perturbation set may include at least a portion of the first plurality of residues in the first perturbation set that form an induced-fit pocket within the first conformation of the protein molecule.

[0014] In some variations, each run of the molecular dynamics simulation can simulate the temporal evolution of the bound protein-structure-compound, which includes a compound bound to the first conformation of the protein molecule at a plurality of first residues included in the first perturbation set.

[0015] In some variations, the bound protein-structure-compound can be generated based at least on the first perturbation set and the structural information that specifies the spatial arrangement within the first conformation of the protein molecule in the unbound state.

[0016] In some variations, determining the first perturbation set can further include excluding from the first perturbation set a third residue that requires a threshold amount of reorientation to interact with at least the compound.

[0017] In some variations, the threshold amount of reorientation can be satisfied by (i) the amount of atoms in the third residue that require reorientation to interact with the compound, (ii) the angle of reorientation that one or more atoms in the third residue need to take to interact with the compound, and / or (iii) the distance of reorientation that one or more atoms in the third residue need to move to interact with the compound.

[0018] In some variations, a first vector from the geometric center (COG) of the plurality of first residues within the first perturbation set to the alpha-carbon of the third residue can be determined. A second vector from the alpha-carbon to one of the delta-carbon, gamma-carbon, or beta-carbon present in the third residue can be determined. The threshold amount of reorientation can be determined to be satisfied based at least on the angle formed by the first vector and the second vector that satisfies one or more thresholds.

[0019] In some variations, the third residue can be retained within the first perturbation set based at least on the third residue being glycine.

[0020] In some variations, determining the first perturbation set may further include determining the distance between a first centroid of a first cluster of residues and a second centroid of a second cluster of residues, determining that the distance does not correspond to one or more dimensions of the compound, and excluding a first residue from the first perturbation set based at least on the first residue being part of the first cluster of residues.

[0021] In some variations, determining the first perturbation set may further include applying a clustering technique that partitions a first plurality of residues into at least a first cluster of residues and a second cluster of residues to the first plurality of residues.

[0022] In some variations, the clustering technique may include one or more of k-means, Mean-Shift clustering, DBSCAN (density-based spatial clustering of applications with noise), expectation maximization (EM) clustering using a Gaussian mixture model (GMM), and agglomerative hierarchical clustering.

[0023] In some variations, determining the first perturbation set may further include determining a first amount of residues within the first cluster of residues and a second amount of residues within the second cluster of residues, determining a first average distance between the first centroid and the residues within the first cluster, and determining a second average distance between the second centroid and the residues within the second cluster.

[0024] In some variations, determining the first perturbation set may further include excluding a first cluster of residues from the first perturbation set based on a determination that, at least from the first perturbation set, (i) a second cluster contains fewer residues than the first cluster, and (ii) the distance between a first centroid of the first cluster and a second centroid of the second cluster exceeds the sum of a first average distance between the first centroid and the residues within the first cluster and a second average distance between the second centroid and the residues within the second cluster.

[0025] In some variations, the result may include the distance between a first geometric center of the compound and a second geometric center of a plurality of backbone nitrogen (N) atoms among the first plurality of residues within the first perturbation set.

[0026] In some variations, the result may include the distance between a first geometric center of the compound and one or more amide nitrogen atoms present among the first plurality of residues within the first perturbation set.

[0027] In some variations, the one or more molecular dynamics simulations may include a first molecular dynamics simulation performed at a first temperature and a second molecular dynamics simulation performed at a second temperature.

[0028] In some variations, the one or more molecular dynamics simulations may include a first molecular dynamics simulation performed over a first length of time and a second molecular dynamics simulation performed over a second length of time.

[0029] In some variations, the one or more molecular dynamics simulations may include a first molecular dynamics simulation performed for a first conformation of a protein molecule associated with a first perturbation set and a second molecular dynamics simulation performed for a second conformation of a protein molecule associated with a second perturbation set.

[0030] In some variations, each of the first conformation and the second conformation can be a three-dimensional structure having different spatial arrangements of atoms within the protein molecule.

[0031] In some variations, a second perturbation set of the second conformation of the protein molecule can be determined. Determining can include at least one of: (i) identifying one or more residues to include in the second perturbation set based at least on the shielding factors of at least a portion of the residues in the protein molecule; (ii) excluding from the second perturbation set a third residue for which the distance to at least a fourth residue within the second perturbation set does not correspond to one or more dimensions of the compound; and (iii) excluding from the second perturbation set a fifth residue that requires a threshold amount of reorientation to interact with the compound.

[0032] In some variations, one or more molecular dynamics simulations can include multiple molecular dynamics simulations. Each molecular dynamics simulation can expose a bound compound-protein-structure that includes a compound bound to the protein molecule at a first plurality of residues within a first perturbation set to different conditions.

[0033] In some variations, the results of one or more molecular dynamics simulations can include one or more conditions under which the bound compound-protein-structure dissociates.

[0034] In some variations, one or more conditions can include the temperature at which the bound compound-protein-structure dissociates.

[0035] In some variations, one or more conditions can include the length of time before the bound compound-protein-structure dissociates.

[0036] In some variations, one or more molecular dynamics simulations include one molecular dynamics simulation that is performed at a higher temperature than another molecular dynamics simulation in response to a bound compound - protein - structure that includes a compound bound to a protein molecule at a first plurality of residues within a first perturbation set not dissociating during another threshold amount of molecular dynamics simulations or at the end of the same simulation.

[0037] In some variations, one or more molecular dynamics simulations include one molecular dynamics simulation that is performed at a lower temperature than another molecular dynamics simulation in response to a bound compound - protein - structure that includes a compound bound to a protein molecule at a first plurality of residues within a first perturbation set not dissociating during another threshold amount of molecular dynamics simulations or at the end of the same simulation.

[0038] In some variations, one or more molecular dynamics simulations include one molecular dynamics simulation that is performed over a longer period than another molecular dynamics simulation in response to a bound compound - protein - structure that includes a compound bound to a protein molecule at a first plurality of residues within a first perturbation set not dissociating during another threshold amount of molecular dynamics simulations or at the end of the same simulation.

[0039] In some variations, one or more molecular dynamics simulations include one molecular dynamics simulation that is performed over a shorter period than another molecular dynamics simulation in response to a bound compound - protein - structure that includes a compound bound to a protein molecule at a first plurality of residues within a first perturbation set dissociating during another threshold amount of molecular dynamics simulations or at the end of the same simulation.

[0040] In some variations, the first plurality of residues within the perturbation set are identified as being those that form a stable bond by determining, at least based on the results of one or more molecular dynamics simulations, one or more metrics that quantify the binding affinity between the compound and the first conformation of the protein molecule, and determining, at least based on the one or more metrics, that the first plurality of residues within the perturbation set form a stable bond with the first conformation of the protein molecule.

[0041] In some variations, determining the first perturbation set can further include determining the distance between a first residue and a second residue and excluding the first residue from the first perturbation set in response to determining that the distance between the first residue and the second residue exceeds one or more dimensions of the compound.

[0042] Implementations of the subject matter described herein can include, but are not limited to, methods consistent with the descriptions provided herein, and articles that include a tangible, embodied machine-readable medium operable to cause one or more machines (e.g., computers, etc.) to carry out one or more of the operations implementing one or more of the features described herein. Similarly, a computer system can be described that can include one or more processors and one or more memories coupled to the one or more processors. The memory, which can include a non-transitory computer-readable or machine-readable storage medium, can encode, include, or store one or more programs to cause the one or more processors to carry out one or more of the operations described herein. A computer-implemented method consistent with one or more implementations of the subject matter can be implemented by one or more data processors present in a single computing system or multiple computing systems. Such multiple computing systems can be connected via one or more connections including, for example, connections via a network (e.g., the Internet, a wireless wide area network, a local area network, a wide area network, or a wired network, etc.) such as a direct connection between one or more of the multiple computing systems, and can exchange data and / or commands or other instructions, etc.

[0043] Details of one or more variations of the subject matter described herein are set forth in the accompanying drawings and the description below. Other features and advantages of the subject matter described herein will be apparent from the description, the drawings, and the claims. Some of the features disclosed herein are described for purposes of illustration in identifying binding sites in protein molecules, but it should be readily understood that such features are not limiting. The following claims are intended to define the scope of the protected subject matter.

Brief Description of the Drawings

[0044] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate some aspects of the subject matter disclosed herein and, together with the description, serve to explain some of the principles associated with the means disclosed herein. At least one drawing is created in color in the patent or application file. A copy of this patent or patent application publication that includes a color drawing will be provided by the Patent Office upon request and payment of the required fee.

[0045]

Figure 1

Figure 2A

Figure 2B

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

[0046] In practice, similar reference numerals are used to refer to the same or similar items in the figures.

DETAILED DESCRIPTION OF THE INVENTION

[0047] The interaction between two molecules, such as the binding between a compound (e.g., a ligand) and a protein molecule, can play an important role in biological processes. Therefore, understanding the dynamics of one or more protein-compound interactions, including the identification of potential binding sites on the protein molecule for binding to the compound, can be desirable for various scientific endeavors (e.g., drug discovery). The binding affinity between a protein molecule and a compound can depend on whether the three-dimensional structure of the protein molecule (e.g., secondary structure, and / or tertiary structure, etc.) complements the same structure of the compound. However, due to its flexibility, the three-dimensional structure of a protein molecule can change over time, especially when the protein molecule approaches and interacts with the compound. Furthermore, in some cases, a population of protein molecules having the same primary structure (e.g., the sequence of amino acid residues) can also exhibit different three-dimensional structures. The terms "protein conformation", "conformation", and "conformer" can be used interchangeably herein to refer to a specific three-dimensional structure of a protein molecule. The same sequence of amino acid residues can be associated with a conformational ensemble (CE) that includes multiple conformations (conformers), each defined by the spatial arrangement of atoms in the constituent amino acid residues.

[0048] Existing techniques that can identify the energetically favorable portions of a protein molecule for binding to a compound ignore the dynamic nature of the protein molecule during the binding process, including the changes in the three-dimensional structure of the protein molecule that occur when the protein molecule is in proximity to and interacting with the compound. As a result, conventional techniques are inaccurate when the structural dynamics of the protein molecule affect the binding between the protein molecule and the compound. Other existing techniques rely on trial and error to calculate the binding energy between a compound and a protein molecule at multiple positions on a specific conformation of the protein molecule before the position with the minimum binding energy is identified. However, a purely trial-and-error-based approach consumes excessive computational resources and imposes a large runtime, making it too inefficient for practical use.

[0049] The interaction between a compound and a protein molecule (e.g., during the binding process) may change the three-dimensional structure of the protein molecule. For example, the presence of a compound can cause reorientation of one or more amino acid residues in the protein molecule, thereby affecting its three-dimensional structure. Such changes can affect the ability of at least some residues to form a stable bond between the compound and the protein molecule. For example, one or more residues in the protein molecule can reorient to form an induced-fit pocket, in which case the residues forming the pocket are more likely to form a stable bond with the compound. However, as already mentioned, existing techniques for identifying binding sites in protein molecules ignore the dynamic nature of protein molecules, including those caused by the interaction between the protein molecule and the compound.

[0050] In contrast, according to some implementations of the subject matter of the present invention, an analysis controller may identify a perturbation set that includes one or more residues within a particular conformation of a protein molecule that can form a stable bond with a compound. The analysis controller may identify the perturbation set for each conformation of the protein molecule in an iterative manner to account for the structural changes that occur in the protein molecule when it interacts with and binds to the compound. Further, it should be understood that the perturbation set of conformations of the protein does not necessarily include all residues that will interact with and bind to the compound. That is, the residues included in the perturbation set are more likely to form a bond between the compound and the corresponding conformation of the protein, but other residues in the protein molecule that are not identified as candidates for inclusion in the perturbation set may also interact with the compound as part of the binding process.

[0051] In some exemplary embodiments, the analysis controller may generate a separate perturbation set for each conformation of the protein molecule. For example, the analysis controller may identify one or more residues within that conformation to include in the perturbation set for each conformation of the protein molecule. In some cases, one or more residues may be identified as being included in the perturbation set of the protein conformation based on the protection factor associated with the individual residues within the conformation of the protein. Further, in some cases, the perturbation set of the protein conformation may be further updated to exclude one or more residues based on geometric properties of the residues, such as the relative position and / or orientation of the residues within that particular conformation of the protein molecule. The analysis controller may determine structural information (e.g., three-dimensional coordinates, etc.) of the bound protein-structure-compound formed by the compound bound to the conformation of the protein molecule at the residues included in the perturbation set prior to performing one or more molecular dynamics (MD) simulations on the bound protein-structure-compound. For example, each molecular dynamics simulation may include a simulation operation of dissociating the bound protein-structure-compound at a different temperature. In some cases, one molecular dynamics simulation may be performed at a different temperature than another (e.g., the previous) molecular dynamics simulation. Alternatively, and / or additionally, one molecular dynamics simulation may be performed at the same temperature as another (e.g., the previous) molecular dynamics simulation, but for a further period of time. The analysis controller may determine whether the residues within the perturbation set form a stable bond between the conformation of the protein and the compound, based at least on the results of the molecular dynamics simulation.

[0052] As already described, in some exemplary embodiments, one or more residues within a particular protein conformation can be selected to be included in the corresponding perturbation set, based at least on the protection factor associated with the individual residues within the protein conformation. The protection factor of a residue can indicate the ability of the residue to interact with a compound, e.g., the ability of the residue to bind to a compound. For example, in some cases, the protection factor of a residue can correspond to the difference between a first energy of the residue in the bound state and a second energy of the residue in the unbound state. In some cases, the protection factor of a residue can be determined based on the amide hydrogen exchange (HX) rate of the protein conformation, and this value can be determined experimentally, e.g., through mass spectrometry, and / or nuclear magnetic resonance, etc. Residues exhibiting a higher protection factor have a higher a priori likelihood of binding to a compound than residues exhibiting a lower protection factor. Thus, when the protection factor of one or more residues meets one or more thresholds, the analysis controller can identify one or more residues in the protein molecule to be included in the perturbation set. In so doing, the analysis controller can identify one or more residues within the conformation of the protein molecule that have a likelihood of forming a stable bond with a compound that meets one or more thresholds for inclusion in the perturbation set.

[0053] In some cases, the ability of two separate residues within a particular conformation of a protein molecule to simultaneously bind to a compound may depend on their spatial arrangement within that conformation of the protein molecule, such as their relative positions. For example, if the distance between a first residue and a second residue (e.g., Euclidean distance, etc.) exceeds one or more dimensions of the compound (e.g., greater than the largest dimension associated with the compound), at least the compound cannot simultaneously bind to both the first residue and the second residue, so at least one of the residues may be excluded from the perturbation set. Thus, in some cases, the analysis controller may determine a perturbation set of the protein conformation by excluding one or more residues based at least on the positions of the residues within that conformation of the protein molecule. For example, if it is determined that the compound cannot simultaneously bind to both the first residue and the second residue, the analysis controller may be able to include the first residue in the perturbation set, but not the second residue. As described in more detail below, in some cases, the analysis controller may further refine the perturbation set using the first residue by excluding one or more residues that cannot form a stable bond with the compound due at least to their orientation. Further, it should be understood that the perturbation set of a particular protein conformation may be reorganized based on the results of molecular dynamics simulations (MD) using one or more modifications to its constituent residues. For example, the perturbation set may be reorganized when the protein molecule and the compound dissociate after one or more molecular dynamics simulations (MD), and then one or more additional molecular dynamics simulations (MD) are performed based on the reorganized perturbation set.

[0054] In some cases, two or more clusters of residues may not be able to bind simultaneously with a compound due to the spatial arrangement of two or more clusters of residues within the conformation of a particular protein. For example, if the distance (e.g., Euclidean distance, etc.) between a first cluster of residues and a second cluster of residues exceeds one or more dimensions of the compound (e.g., is greater than the largest dimension of the compound), the compound cannot bind to both clusters simultaneously, and thus at least one cluster of residues may be excluded from the perturbation set. In some cases, the analysis controller may identify one or more clusters of residues based on the relative positions (e.g., three-dimensional coordinates, etc.) of individual residues within the conformation of the protein. Further, in some cases, the analysis controller may identify one or more clusters of residues by applying clustering techniques including, at least, for example, k-means, Mean-Shift clustering, DBSCAN (density-based spatial clustering of applications with noise), expectation maximization (EM) clustering using a Gaussian mixture model (GMM), and / or agglomerative hierarchical clustering. As described in further detail below, in some cases, the analysis controller may further refine the perturbation set using a first cluster of residues by excluding one or more residues that, at least due to their orientation, cannot form a stable bond with the compound. Further, in some cases, if the protein molecule and the compound dissociate after one or more molecular dynamics simulations (MD), the perturbation set of the conformation of the protein molecule may be reorganized such that one or more additional molecular dynamics simulations (MD) can be performed based on the reorganized perturbation set.

[0055] As already described, in some exemplary embodiments, the analysis controller can determine a perturbation set by excluding one or more residues from at least the perturbation set based at least on the orientation of residues within a particular conformation of a protein molecule. For example, in some cases, since the orientation of the residue reduces the likelihood of a bond between the residue and the compound, the analysis controller can exclude the residue. Excluding a residue from the perturbation set corresponds to the observation that the orientation of the residue with respect to the compound (e.g., when the compound approaches the protein molecule) may have a higher energy threshold for the residue to bind to the compound if the residue requires a significant reorientation to bind to the compound. Thus, in some cases, the analysis controller can exclude a residue when the orientation of the residue within the conformation of the protein requires a threshold amount of reorientation for the residue to bind to the compound. In some cases, such a threshold amount of reorientation can be satisfied, for example, by reorientation of a threshold amount of atoms in the residue (e.g., a threshold amount of side chain atoms and / or backbone atoms). Alternatively, and / or additionally, such a threshold amount of reorientation can be satisfied by the angle and / or distance of reorientation of one or more atoms in the residue.

[0056] In some exemplary embodiments, the analysis controller can determine the binding stability provided by residues within the perturbation set based at least on the protection factors, spatial arrangements, and / or orientations of individual residues (or clusters of residues) within the corresponding conformations of the protein molecules. For example, in some cases, the binding stability provided by residues within the perturbation set can be determined by performing one or more molecular dynamics simulations (MD) based at least on the bound protein-structure-compound in which the compound is bound to the conformation of the protein molecule at the residues included in the perturbation set. The bound protein-structure-compound can be determined by applying one or more docking algorithms to generate composite structure information of the bound protein-structure-compound based on the perturbation set of the protein conformation and the unbound structure information. In some cases, the molecular dynamics simulation can be performed based on the composite structure information of the bound protein-structure-compound shown, for example, in the composite coordinate file associated with the bound protein-structure-compound.

[0057] In some exemplary embodiments, the analysis controller may perform multiple iterative molecular dynamics (MD) simulations where each molecular dynamics simulation exposes the bound protein-structure-compound to different conditions, for example, until the bound protein-structure-compound dissociates. For example, in some cases, the first molecular dynamics simulation may be performed at a different temperature than the second molecular dynamics simulation to simulate the time evolution of the bound protein-structure-compound at multiple temperatures. Alternatively and / or additionally, the second molecular dynamics simulation may be performed at the same temperature as the first molecular dynamics simulation but for an additional period of time. In some cases, if the bound protein-structure-compound does not dissociate after one or more molecular dynamics simulations performed at a first temperature, the analysis controller may, in some cases, perform one or more additional molecular dynamics simulations at a second temperature that may be higher than the first temperature. The analysis controller may continue to perform additional molecular dynamics simulations on the bound protein-structure compound by increasing the temperature incrementally and / or extending the time incrementally until the bound protein-structure-compound dissociates. In so doing, the analysis controller may determine and / or rank the stability of the bonds formed by the residues within the perturbation set, at least by determining the temperature at which the bound protein-structure-compound dissociates and / or the duration of the bound protein-structure-compound (e.g., the length of time until the bound protein-structure-compound dissociates). For example, if it is determined that the bound protein-structure-compound dissociates at a higher temperature and / or after a longer period of time, the analysis controller may determine that the stability of the bond between the compound and the conformation of the protein at the residues included in the perturbation set is greater.

[0058] In some exemplary embodiments, the analysis controller may perform multiple molecular dynamics simulations (MD) for different conformations of a protein molecule to account for fluctuations in the three-dimensional structure of the same protein molecule that occur over time and are observed in a population of the same molecules due to folding of the protein molecule. For example, in some cases, the analysis controller may perform a first molecular dynamics simulation for a first conformation of the protein molecule and a second molecular dynamics simulation for a second conformer of the protein molecule. Here, the analysis controller may receive structure information (e.g., an unbound coordinate file) indicating, for each conformation of the protein molecule, the spatial arrangement of various atoms in the protein molecule (e.g., the spatial positions of individual atoms relative to each other). Further, each conformation of the protein molecule may be associated with a different perturbation set since changes within the spatial three-dimensional structure of the protein molecule can affect corresponding changes in residues that are more likely to form a stable bond with a compound. Thus, the analysis controller can generate a first perturbation set for a first conformation of the protein molecule and a second perturbation set for a second conformation of the protein molecule. Each of the first perturbation set and the second perturbation set may include residues identified based on, for example, the shielding factor, spatial arrangement, and / or orientation of individual residues (or clusters of residues) within the corresponding conformation of the protein molecule. The analysis controller can determine the relative binding affinity and / or ranking between a compound and the protein molecule, as indicated by the binding affinity between the compound and the protein molecule over multiple conformations of the protein molecule, based at least on the results of the molecular dynamics simulations (MD).

[0059] FIG. 1 is a system diagram showing an example of a protein structure analysis system 100 according to some exemplary embodiments. As shown in FIG. 1, the protein structure analysis system 100 may include an analysis controller 110 communicatively coupled to a client device 130 via a network 140. The network 140 may be a wired network and / or a wireless network including, for example, a local area network (LAN), a virtual local area network (VLAN), a wide area network (WAN), a public land mobile network (PLMN), and / or the Internet. The client device 130 may be a processor-based device including, for example, a workstation, a desktop computer, a laptop computer, a high-performance computer, a smartphone, a tablet computer, and / or a wearable device.

[0060] In some exemplary embodiments, the analysis controller 110 may identify one or more residues in a protein molecule that can form a stable bond with a compound (e.g., a ligand, etc.). As shown in FIG. 1, in some cases, the analysis controller 110 may include a perturbation set engine 112 that identifies one or more residues that can form a stable bond between the corresponding conformation of the protein molecule and the compound as part of each perturbation set associated with the protein molecule. For example, in some cases, one or more residues that can provide a stable bond with a compound may be identified based on the protection factor, spatial arrangement, and / or orientation of individual residues within the corresponding conformation of the protein molecule. Further, in the embodiment shown in FIG. 1, the analysis controller 110 may include a molecular dynamics (MD) simulation engine 114 that performs one or more molecular dynamics simulations.

[0061] In some cases, the molecular dynamics simulation engine 114 can perform one or more molecular dynamics simulations for each perturbation set, with each run of the molecular dynamics simulation being performed at a different temperature or, in some cases, at the same temperature but for an additional period of time. For example, the molecular dynamics simulation engine 114 can perform a first molecular dynamics simulation 115a at a first temperature and a second molecular dynamics simulation 115b at the same first temperature or a different second temperature for a first perturbation set 113a of a first conformation 119a of a protein molecule. As shown in FIG. 1, the analysis controller 110 can include an evaluation engine 118 that determines, based at least on the first molecular dynamics simulation 115a and the second molecular dynamics simulation 115b, whether the residues included in the first perturbation set 113a form a stable bond between the compound and the first conformation 119a of the protein molecule.

[0062] As will be described in more detail below, in cases where the perturbation set engine 112 generates a plurality of perturbation sets including separate perturbation sets for each conformation of the protein molecule, the molecular dynamics simulation engine 114 can perform one or more molecular dynamics simulations (e.g., at different temperatures and / or over different lengths of time) for each perturbation set. For example, the molecular dynamics simulation engine 114 can perform a first molecular dynamics simulation 115 for a first perturbation set 113a of a first conformation 119a and a second molecular dynamics simulation 115b for a second perturbation set 113b of a second conformation 119b. In such cases, the evaluation engine 118 can further determine, based on the results of at least the first molecular dynamics simulation 115a and the second molecular dynamics simulation 115b, for example, the stability of the bond between the protein molecule and the compound over a plurality of conformations of the protein molecule.

[0063] More specifically, FIG. 2A is a flow diagram showing an example of a method 200 for identifying residues within a particular conformation of a protein molecule that can form a stable bond with a compound and a protein molecule according to some exemplary embodiments. As shown in FIGS. 1 and 2A, method 200 can be performed by analysis controller 110, for example, to determine whether residues included in a perturbation set of a protein conformation form a stable bond between the protein molecule and a compound (e.g., a ligand, etc.).

[0064] At 202, the analysis controller 110 may determine a perturbation set of the conformation of the protein molecule. In some exemplary embodiments, the analysis controller 110, for example, the perturbation set engine 112, may generate a first perturbation set 113a including one or more residues from a first conformation 119a of the protein molecule. In some cases, the perturbation set engine 112 may generate the first perturbation set 113a by identifying, at least based on the shielding factor, spatial arrangement, and / or orientation of individual residues within at least the first conformation 119a of the protein molecule, one or more residues to include in the first perturbation set 113a. For example, in some cases, the perturbation set engine 112 may determine to include a first residue in the first perturbation set 113a based on the shielding factor of the first residue satisfying at least one or more thresholds. In the embodiment shown in FIG. 1, the shielding factor of the first residue may be included in first shielding factor data 117a received from a shielding factor analyzer 116 (such as a mass spectrometer and / or a nuclear magnetic resonance (NMR) device, etc.). In some cases, the shielding factor of the first residue may satisfy one or more thresholds if its value is within a particular percentile (or quantile) of the shielding factor values observed in the first conformation 119a of the protein molecule. For example, in some cases, the shielding factor of the first residue may satisfy one or more thresholds if its value is within X% of the maximum shielding factor value (or maximum log(PF)), where X is a value between 5 and 20. In that case, if the maximum log(PF) is 2, the shielding factor of the first residue will be within the 20th percentile if its value exceeds 1.6. Alternatively, if the maximum log(PF) is 4, the shielding factor of the first residue will be within the 20th percentile if its value exceeds 3.2. It should be understood that the value of the percentile X can be determined, and in some cases adjusted, based on the magnitude of the difference between the maximum shielding factor value observed in the first conformation 119a of the protein molecule and the other shielding factor values of other residues within the first conformation 119a of the protein molecule.

[0065] However, in some cases, if the perturbation set engine 112 determines that the distance (e.g., Euclidean distance) between a first residue (or a first cluster of residues including the first residue) and a second residue (or a second cluster of residues including the second residue) within the first conformation 119a of the protein molecule exceeds one or more dimensions (e.g., the largest dimension) associated with the compound, the perturbation set engine 112 may subsequently determine to exclude the first residue from the first perturbation set 113a. As described in more detail below, the residues within the first conformation 119a of the protein molecule can be grouped into one or more clusters, and in some cases, for example, the distance (e.g., Euclidean distance) between the first centroid of the first cluster and the second centroid of the second cluster exceeds one or more dimensions of the compound, and if the compound cannot bind to both the residues within the first cluster and the residues within the second cluster simultaneously, the first cluster of residues can be excluded from the first perturbation set 113a. The exclusion of the residues within the first cluster reflects the observation that a compound bound to the residues within the second cluster cannot simultaneously bind to the residues within the first cluster either.

[0066] In some cases, the perturbation set engine 112 can further determine the first perturbation set 113a by excluding at least one or more residues from the first perturbation set 113a based at least on the orientation of one or more residues. As described in more detail with reference to FIG. 10, the orientation of a residue is based on the alpha-carbon (c α ) in the residue, the center of geometry (COG) of the perturbation set containing the residue, and the beta-carbon (c β ) in the residue, the delta-carbon (c δ ), or the gamma-carbon (c γ) can be determined based on one or more positions among them. In some cases, if a residue requires a threshold amount of reorientation to bind to a compound, the residue can be excluded from the first perturbation set 113a. For example, in some cases, if the amount of atoms in a residue that requires reorientation meets one or more thresholds, such a threshold amount of reorientation can be met. Alternatively, and / or additionally, if the angle and / or distance of reorientation of one or more atoms in a residue meets one or more thresholds, the threshold amount of reorientation can be met. In some cases, if the side chain of a residue requires a threshold amount of reorientation to bind to a compound, the residue can be excluded from the first perturbation set 113a. In these cases, the first vector (defined by the first position of the geometric center of the perturbation set containing the residue and the α-carbon (c α ) and the second position) and the second vector (the second position of the α-carbon (c α ) and the third position of the β-carbon (c β ), δ-carbon (c δ ), or γ-carbon (c γ ) defined by) meet one or more thresholds, the threshold amount of reorientation can be met.

[0067] In 204, the analysis controller 110 can perform one or more molecular dynamics simulations to generate results indicating the interactions between a compound and a plurality of residues within a perturbation set. In some exemplary embodiments, the molecular dynamics (MD) simulation engine 114 can perform one or more molecular dynamics (MD) simulations based at least on, for example, the generated first perturbation set 113 of operation 202 and the structural information that describes the spatial arrangement of atoms within the first conformation 119a of the protein molecule. For example, in some cases, the structural information associated with the first conformation 119a of the protein molecule can indicate the positions (e.g., coordinates) of atoms in the protein molecule. In some cases, in addition to the structural information of the first conformation 119a of the protein molecule, the molecular dynamics simulation engine 114 can receive additional information, such as the positions of one or more heavy atoms (e.g., oxygen, nitrogen, and / or carbon) within the first conformation 119a of the protein molecule, the experimental pH used to define the ionization state on the side chains of the protein molecule, and / or the force field generated for a compound (e.g., a ligand).

[0068] In some exemplary embodiments, the analysis controller 110 can generate a bound compound-protein-structure (e.g., a protein-ligand complex) based at least on a first perturbation set 113a of a first conformation 119a of a protein molecule and structural information. For example, in some cases, the analysis controller 110 can apply a docking algorithm to generate a bound compound-protein-structure when the compound is bound to the protein molecule at residues included in the first perturbation set. Further, in some cases, the analysis controller 110 can generate composite structural information of the bound compound-protein-structure based at least on the first perturbation set 113a of the first conformation 119a of the protein molecule and unbound structural information. In some cases, the molecular dynamics simulation engine 114 can perform one or more molecular dynamics simulations based at least on the composite structural information of the bound compound-protein-structure. For example, the molecular dynamics simulation engine 114 can perform a first molecular dynamics simulation 115a at a first temperature and a second molecular dynamics simulation 115b for a further length of time at the same first temperature or at a second temperature (different from the first temperature). Iterative rounds of molecular dynamics simulations (MD) are described in further detail in FIGS. 2B, 11, and 12.

[0069] In 206, the analysis controller 110 can identify a plurality of residues within the perturbation set as those that form a stable bond between the compound and the conformation of the protein molecule, based at least on the results of one or more molecular dynamics simulations. In some exemplary embodiments, the results of one or more molecular dynamics simulations, such as forming a first molecular dynamics simulation 113a and a second molecular dynamics simulation 113b, can include the temperature at which the bound compound-protein-structure dissociates and / or the duration of the bound compound-protein-structure (e.g., how long the bound compound-protein-structure remained bound). In some cases, the evaluation engine 118 can determine one or more measurement criteria for quantifying the binding affinity (e.g., an achieved score, and / or a deviation score, etc.) between the compound and the first conformation 119a of the protein molecule, based at least on the results of the molecular dynamics simulation. Further, in some cases, the evaluation engine 118 can identify a residue within the perturbation set as one that forms a stable bond between the compound and the first conformation 119a of the protein molecule, based at least on one or more measurement criteria that meet one or more thresholds. The analysis performed to identify the perturbation set as one that forms a stable bond between the compound and a particular conformation of the protein molecule is described in further detail in FIGS. 11 and 12.

[0070] Figure 2B is a flow diagram illustrating another example of a method for determining one or more binding conformations of protein molecules bound to a compound, according to some exemplary embodiments. As shown in FIGS. 1 and 2A, method 250 can be implemented by analysis controller 110 to determine the binding affinity between a protein molecule and a compound (e.g., a ligand, etc.) based on the binding stability between the compound and multiple conformations of the protein molecule. As already mentioned, the folding of a protein molecule can cause changes in the three-dimensional structure of the protein molecule. For example, the protein molecule can assume a first conformation 119a at a first time point and a second conformation 119b at a second time point, respectively. Further, at any given time, a population of the same protein molecules can include multiple conformations of the protein molecule, including, for example, the first conformation 119a and the second conformation 119b. Changes in the conformation of a protein molecule can alter the residues in the protein molecule that can form a stable bond with a compound. Thus, in some exemplary embodiments, analysis controller 110 can determine multiple sets of perturbations, each set of perturbations being associated with a different conformation of the protein molecule and including one or more residues in the protein molecule that can form a stable bond with the compound while the protein molecule assumes the corresponding three-dimensional structure. As described in further detail below, method 250 can be implemented to account for the structural dynamics of the protein molecule, including the temporal evolution of its three-dimensional structure and the associated fluctuations in the interaction between the protein molecule and the compound.

[0071] At 252, the analysis controller 110 may determine a first set of perturbations of a first conformation of a protein molecule. In some exemplary embodiments, the analysis controller 110, e.g., the perturbation set engine 112, may determine a first set of perturbations 113a of a first conformation of a protein molecule. In the embodiment shown in FIG. 1, the perturbation set engine 112 may determine the first set of perturbations 113a based at least on first protection factor data 117a (e.g., received from the protection factor analyzer 116) associated with the first conformation 119a of the protein molecule. Alternatively and / or additionally, the perturbation set engine 112 may determine the first set of perturbations 113a based on the spatial arrangement and / or orientation of atoms in the individual residues that form the first conformation 119a of the protein molecule.

[0072] At 254, the analysis controller 110 may determine a second set of perturbations of a second conformation of a protein molecule. In some exemplary embodiments, the analysis controller 110, e.g., the perturbation set engine 112, may determine a second set of perturbations 113b of a second conformation of a protein molecule. For example, in the embodiment shown in FIG. 1, the perturbation set engine 112 may determine the second set of perturbations 113b based at least on second protection factor data 117b (e.g., received from the protection factor analyzer 116) associated with the second conformation 119b of the protein molecule. Alternatively and / or additionally, the perturbation set engine 112 may determine the second set of perturbations 113b based on the spatial arrangement and / or orientation of atoms in the individual residues that form the second conformation 119b of the protein molecule.

[0073] At 256, the analysis controller 110 can perform at least a first molecular dynamics simulation to generate a first result indicating the interaction between the compound and a first plurality of residues within the first perturbation set. In some exemplary embodiments, the analysis controller 110, for example, the molecular dynamics simulation engine 114, can perform one or more molecular dynamics simulations (MD), such as a first molecular dynamics simulation 115a, on a first bound compound - protein - structure corresponding to a compound bound to a first conformation 119a of the protein molecule at residues included in the first perturbation set 113a. In some cases, the molecular dynamics analysis engine 114 can perform multiple molecular dynamics simulations on the first bound compound - protein - structure, for example, at different temperatures and / or over different durations. The multiple molecular dynamics simulations can be performed, for example, to determine a first temperature at which the first bound compound - protein - structure dissociates and / or a first duration of the first bound compound - protein - structure. At that time, the analysis controller 110, for example, the evaluation engine 118, can determine whether the first conformation 119a of the protein molecule can form a stable bond with the compound at the residues included in the first perturbation set 113a.

[0074] In 258, the analysis controller 110 can perform at least a second molecular dynamics simulation to generate a second result indicating the interaction between the compound and a second plurality of residues within the second perturbation set. In some exemplary embodiments, the analysis controller 110, e.g., the molecular dynamics simulation engine 114, can perform one or more molecular dynamics (MD) simulations, e.g., a second molecular dynamics simulation 115b, on a second bound compound-protein-structure corresponding to the compound bound to the second conformation 119b of the protein molecule at the residues included in the second perturbation set 113b. For example, in some cases, the molecular dynamics analysis engine 114 can perform multiple molecular dynamics simulations on the second bound compound-protein-structure at different temperatures and / or over different durations. Further, in some cases, the multiple molecular dynamics simulations can be performed, e.g., to determine a second temperature at which the second bound compound-protein-structure dissociates and / or a second duration of the second bound compound-protein-structure. At that time, the analysis controller 110, e.g., the evaluation engine 118, can determine whether the second conformation 119b of the protein molecule can form a stable bond with the compound at the residues included in the second perturbation set 113b.

[0075] In 260, the analysis controller 110 can determine the binding affinity between a compound and a protein molecule based on at least the first result and the second result. In some exemplary embodiments, the analysis controller 110, for example, the evaluation engine 118, can determine the binding affinity between a compound and multiple conformations of a protein molecule based on at least the results of molecular dynamics simulations performed on different conformations of the protein molecule. As already described, the evaluation engine 118 can determine the stability of the binding between the compound and the protein molecule within the first conformation 119a (e.g., operation 256) and the stability of the binding between the compound and the protein molecule within the second conformation 119b (e.g., operation 258). Thereby, the structural dynamics of the protein molecule, including its time-dependent changes within its three-dimensional structure, can be explained.

[0076] Accordingly, in some cases, the evaluation engine 118 may further determine one or more metrics (e.g., an achievement score, and / or a deviation score, etc.) for quantifying the binding affinity between a compound and a protein molecule based on the binding stability between the compound and various conformations of the protein molecule. For example, in some cases, the evaluation engine 118 may determine that the binding affinity between a compound and a protein molecule meets one or more thresholds if the compound can form a stable bond with a threshold amount of conformations of the protein molecule. Alternatively, and / or additionally, the evaluation engine 118 may identify one or more of the first conformation 119a and the second conformation 119b as the stable docking conformation (or the most stable conformation) of the protein molecule based at least on the results of the first molecular dynamics simulation 115a performed on the first conformation 119a and the results of the second molecular dynamics simulation 115b performed on the second conformation 119b. If the compound can form stable bonds with multiple conformations of the protein molecule, the evaluation engine 118 may determine that the multiple conformations of the protein molecule can stably bind to the compound.

[0077] Figure 3 is a flow diagram illustrating one example of a method 300 for determining a perturbation set according to some exemplary embodiments. As shown in FIGS. 1, 2A - B, and 3, the method 300 can be performed by an analysis controller 110, e.g., a perturbation set engine 112, to determine a perturbation set of a particular conformation of a protein molecule. In some cases, the method 300 may implement operation 202 of method 200, as well as operations 252 and 254 of method 250.

[0078] At 302, the analysis controller 110 can identify one or more residues to include in the perturbation set, at least based on the protection factor of each residue within the conformation of the protein molecule. In some exemplary embodiments, the analysis controller 110, for example, the perturbation set engine 112, can generate a perturbation set, such as a first perturbation set 113a or a second perturbation set 113b, by identifying one or more residues within one or more regions of the corresponding protein conformation that exhibit a protection factor satisfying at least one or more thresholds. For example, in some cases, the perturbation set can be determined by including one or more residues from the protein molecule that exhibit a protection factor exceeding the threshold within the perturbation set. FIG. 4 shows an example of a protein conformation 402 and a protection factor diagram 404 showing the values of the protection factors associated with each residue within the protein conformation 402. The protection factor diagram 404 includes the values of the protection factors as a function of the residues within the protein conformation 402. As already described, in some cases, one or more residues whose protection factor satisfies one or more thresholds can be included in the perturbation set. In some cases, the one or more thresholds can include a minimum protection factor value, and / or a maximum protection factor value, etc. Alternatively, the one or more thresholds can include a percentage (e.g., 70% of the maximum protection factor value) of the maximum protection factor value of the residues in the protein molecule. In some cases, the perturbation set engine 112 can generate the perturbation set by including a threshold amount of residues (e.g., a maximum amount of residues, and / or a minimum amount of residues, etc.) in the perturbation set. It should be understood that the amount of residues included in the perturbation set containing the aforementioned threshold amount of residues can be fixed or variable for each run of the molecular dynamics simulation (MD). In the example shown in FIG. 4, three residues (numbered 1, 2, and 3) within the protein conformation 402 are identified as being associated with a protection factor exceeding the threshold and are thus identified to be included in the perturbation set by the perturbation set engine 112.

[0079] In 304, the analysis controller 110 may exclude one or more residues from the perturbation set based at least on the spatial arrangement of the residues within the perturbation set. In some exemplary embodiments, determining the perturbation set may include excluding one or more residues from the perturbation set based on the spatial arrangement of one or more residues within the conformation of the protein. For example, the analysis controller 110, e.g., the perturbation set engine 112, may exclude a first residue if the distance between the first residue and a second residue within the conformation of the protein exceeds one or more dimensions of the compound. In some cases, e.g., the perturbation set engine 112 may exclude the first residue if the Euclidean distance between the two closest atoms on the first residue and the second residue exceeds the maximum dimension of the compound.

[0080] In some cases, if the compound bound to the conformation of the protein dissociates in the residues included in the perturbation set, e.g., after one or more molecular dynamics simulations (MD), the perturbation set engine 112 may reconstruct the perturbation set to include the second residue but not the first residue. In some cases, the analysis controller 110 may further refine the perturbation set (and / or the reorganized perturbation set using the second residue) using the first residue by excluding at least one or more residues whose orientation prevents the formation of a stable bond with the compound. Further, in some cases, one or more subsequent rounds of molecular dynamics simulations may be performed to evaluate the stability of the bonds formed by the residues within each perturbation set. At that time, the perturbation set engine 112 may generate a perturbation set, which in some cases may be a reorganization based on the results of molecular dynamics simulations (MD) for each conformation of the protein molecule.

[0081] FIG. 5 shows an example of a compound 502 and a protein molecule 402 that includes three residues (numbered 1, 2, and 3) identified as being included in the perturbation (e.g., operation 302). The compound 502 is characterized by lengths L1, L2, and L3 along three axes (e.g., coordinate axes). If the lengths L1, L2, and L3 (or the longest of the lengths L1, L2, and L3) are less than, for example, the distance (e.g., Euclidean distance) between residue cluster 1 and residue cluster 3, the compound 502 cannot bind to the residues of both residue cluster 1 and residue cluster 3 simultaneously. Similarly, if the lengths L1, L2, and L3 (or the longest of the lengths L1, L2, and L3) are less than the distance (e.g., Euclidean distance) between residue cluster 2 and 3, the compound 502 cannot bind to the residues of both residue cluster 2 and residue cluster 3 simultaneously. Such observations may indicate that a single perturbation set cannot include the residues of both residue cluster 1 and residue cluster 3, or the residues of both residue cluster 2 and residue cluster 3. In contrast, if the distance (e.g., Euclidean distance) between residue cluster 1 and residue cluster 2 is less than at least one of the lengths L1, L2, or L3, all of the residues within residue cluster 1 and residue cluster 2 can be included in the perturbation set.

[0082] In some cases, the analysis controller 110 instructs the perturbation set engine 112 to identify clusters of residues to include in and / or exclude from the perturbation set, in addition to or instead of excluding residues based on the dimensions of the compound (in some cases, additional constants that allow for protein flexibility). For example, in some cases, residues in a protein molecule can be grouped into clusters by applying clustering techniques, e.g., before one or more clusters of residues are excluded from the perturbation set. In addition to or instead of excluding individual residues from the perturbation set, the perturbation set engine 112 can identify clusters of residues to exclude from the perturbation set to account for the vicinity of neighboring residues that can and / or cannot bind to the compound due to the physical dimensions of the compound. As described in further detail below, a first cluster of residues can be identified as being excluded based on the amount and location of the residues included in the cluster, as well as the centroid of the cluster.

[0083] Examples of clustering techniques can include one or more of k-means, Mean-Shift clustering, DBSCAN (density-based spatial clustering of applications with noise), expectation maximization (EM) clustering using a Gaussian mixture model (GMM), and agglomerative hierarchical clustering. FIG. 6A shows the residues within the conformation of a protein prior to clustering, and FIG. 6B shows the residues within the conformation of the protein after clustering. As shown in FIG. 6A, the residues within the conformation of the protein are represented by individual spheres. The result of applying the clustering technique, shown in FIG. 6(B), includes a first cluster 612 of residues indicated by spheres numbered "1" and a second cluster 614 of residues indicated by spheres numbered "2". FIG. 7 is a schematic diagram of the first cluster 612 and the second cluster 614. The first cluster 612 has a first characteristic position 702, and the second cluster 614 has a second characteristic position 704. The first characteristic position 702 and the second characteristic position 704 can be the centroids of the first cluster 612 and the second cluster 614, respectively. The first cluster 612 can be characterized by a first average distance d1, and the second cluster 614 can be characterized by a second average distance d2, where TIFF2025519200000002.tif29170, where n represents the amount of residues within the first cluster 612 and m represents the amount of residues within the second cluster 614; TIFF2025519200000003.tif5170 represents the distance between the i-th residue within the first cluster 612 and the centroid 702 of the first cluster 612; It is the distance between the i-th residue in the second cluster 614 and the centroid 704 of the first cluster 612. The displacement vector R indicates the displacement of the second centroid 704 from the first centroid 702. The distance between the two centroids can be described as the absolute value of the displacement vector R. Based on the foregoing, if the second cluster 614 has fewer residues than the first cluster 612 and the absolute value of the displacement vector R is greater than the sum of the first average distance d1 and the second average distance d2 TIFF2025519200000005.tif5170 The second cluster 614 can be excluded.

[0084] In addition to the exclusion of the second cluster 614, one or more individual residues within the first cluster 612 (e.g., selected for inclusion after the exclusion of the second cluster 614) can be further excluded from the perturbation set based on the orientation of one or more residues within the first cluster 612. For example, in FIG. 8, the residues within the first cluster 612 indicated by the spheres numbered "3" are retained within the perturbation set, and the residues indicated by the spheres numbered "1" are selected to be excluded based on the orientation of these residues. In some cases, the residues indicated by the spheres numbered "1" can be excluded from the perturbation set based at least on residues that require a threshold amount of reorientation to bind to the compound. If the amount of atoms in the residues that require reorientation meets one or more thresholds, such a threshold amount of reorientation can be satisfied. For example, some of the residues indicated by the spheres numbered "1" can contain a threshold amount of atoms that require reorientation to interact with the compound. Alternatively, and / or additionally, some of the residues indicated by the spheres numbered "1" can contain one or more atoms whose reorientation angle and / or reorientation distance meet one or more thresholds. In some cases, some of the residues indicated by the spheres numbered "1" can contain side chains that require a threshold amount of reorientation to bind to the compound. In these cases, the first vector (the first position of the geometric center of the perturbation set containing the residue and the α-carbon (c α) defined by the second position of) and the second vector (alpha-carbon (c in the residue α ) at the second position and beta-carbon (c β ), delta-carbon (c δ ), or gamma-carbon (c γ ) defined by the third position of), when the angle between them satisfies one or more thresholds, a threshold amount of reorientation can be satisfied.

[0085] Returning to Figure 3, at 306, the analysis controller 110 can exclude one or more residues from the perturbation set based on the orientation of at least one or more residues within the perturbation set. In some exemplary embodiments, determining the perturbation set can further include excluding one or more residues (e.g., from the plurality of residues selected in operation 302 and / or from the plurality of residues identified in operation 304, etc.) based on the orientation of at least one or more residues and / or one or more constituent atoms within the one or more residues. For example, in some cases, a residue can be excluded from the perturbation set if its orientation requires a threshold amount of change to interact with the compound. An exemplary calculation for identifying residues to be excluded based on its orientation and / or the orientation of its constituent atoms is shown in FIG. 10. In some cases, when the amount of atoms in a residue that is oriented away from the compound (e.g., oriented at an angle greater than a threshold angle with respect to the compound) satisfies one or more thresholds (e.g., exceeds a specific number), the aforementioned threshold amount of change is satisfied. In this case, since additional energy is required to reorient the residue to bind the compound to the second residue, the residue can be excluded. Thus, the likelihood that the residue forms a stable bond with the compound is lower.

[0086] Figure 9 shows an exemplary portion 900 of the conformation of a protein having a plurality of residues. As shown in Figure 9, the side chains of some residues within the conformation of the protein are oriented towards a first side (side 1), and the side chains of other residues within the conformation of the protein are oriented towards a second side (side 2). When a compound approaches the protein molecule from the first side (side 1), the likelihood of interaction (or binding) between the compound and the residues whose side chains are oriented towards the second side (side 2) of the protein molecule is small. Similarly, when a compound approaches the protein molecule from the second side (side 2), the likelihood of interaction (or binding) between the compound and the residues whose side chains are oriented towards the first side (side 1) is small.

[0087] Figure 10 shows an exemplary method of excluding or including residues based on the orientation of the residues. Figure 10 shows an exemplary residue 1000 having a single alpha-carbon (c α as indicated) located at a particular position (e.g., having spatial coordinates (x,y,z)). Residue 1000 can be part of a perturbation set that includes one or more other residues. Each residue within the perturbation set that includes residue 1000 can be associated with two direction unit vectors (e.g., vectors having a magnitude of 1). For example, if residue 1000 is considered to be TIFF2025519200000006.tif5170, residue 1000 is a first unit vector pointing from the center of geometry (COG) of the perturbation set towards the alpha-carbon (c α ) of residue 1000 TIFF2025519200000007.tif5170 as well as a second unit vector pointing from the alpha-carbon (c α ) of residue 1000 towards the delta-carbon (if a delta-carbon exists), the gamma-carbon (if a gamma-carbon exists), or the beta-carbon (if a beta-carbon exists) of residue 1000 TIFF2025519200000008.tif6170 can be associated with. When residue 1000 is glycine that does not contain a β-carbon, residue 1000 can be included in the perturbation set without being subject to any orientation-based exclusions. Otherwise, if residue 1000 is not glycine, residue 1000 can be retained within the perturbation set if the following restrictions are met: TIFF2025519200000009.tif6170 In some cases, residue 1000 TIFF2025519200000010.tif5170 need not have (such as in the case of glycine) the second unit vector, and in this case too, residue 1000 can be retained within the perturbation set without being subject to orientation-based exclusions.

[0088] In some cases, the geometric center (COG) of the entire perturbation set can be defined by the spatial coordinates (e.g., can be defined by three-dimensional coordinates (x, y, z)) of the backbone nitrogen atom (N) of each residue included in the perturbation set. The geometric center (COG) of the perturbation set can be understood to be defined for each frame of the molecular dynamics simulation (MD). TIFF2025519200000011.tif5170 It should be understood that it can be defined for each frame of the molecular dynamics simulation (MD). Therefore, for the perturbation set of the frames within the trajectory t of the molecular dynamics simulation (MD), TIFF2025519200000012.tif6170 the following equation (wherein, TIFF2025519200000013.tif5170 indicates the vector from the origin of the reference frame to the backbone nitrogen atom of the i-th residue, and r iは is within the perturbation set and N is the amount of residues within the perturbation set) can be determined based on this. TIFF2025519200000014.tif9170

[0089] FIG. 11 is an exemplary schematic diagram of an analysis controller 110, which includes a perturbation set engine 112 and a molecular dynamics simulation engine 114, and a first database 1104 and a second database 1108. The molecular dynamics simulation engine 114 can receive from the first database 1104 at least the first structural information of the first conformation 119a of the protein molecule and the first perturbation set 113a associated with the first conformation 119a generated by the perturbation set engine 112.

[0090] The first database 1104 (also referred to as the unbound protein structure database) can include structural information (e.g., unbound coordinate files) regarding a plurality of conformations (or conformers) of the protein molecule. The structural information regarding the conformation of the protein can indicate the shape of the conformation of the protein, including, for example, by specifying the spatial arrangement of the constituent atoms. As described in more detail below, the various conformations of the protein molecule can be generated by a protein structure information generation algorithm (e.g., an accelerated molecular dynamics analysis algorithm, etc.) that simulates the time evolution of the protein molecule (e.g., the dynamics of the protein) to capture the structural arrangement of the protein molecule (e.g., the spatial arrangement of individual residues and / or constituent atoms) at various time instances in its particular trajectory. The structural arrangement of each individual conformation of the protein and the corresponding perturbation set is provided as an input to the molecular dynamics simulation (MD) engine 114.

[0091] The perturbation set engine 112 can receive the first structural information of the first conformation 119a (e.g., from the first database 1104) and generate a first perturbation set 113a. As already described, the perturbation set engine 112 can identify one or more residues within the first conformation 119a of the protein molecule to be included in the first perturbation set 113a based at least on the protection factors, spatial arrangements, and / or orientations of the residues within the first conformation 119a of the protein molecule (e.g., by implementing the method 300 shown in FIG. 3). In cases where the perturbation set engine 112 generates the first perturbation set 113a by excluding one or more clusters of residues within the first conformation 119a of the protein molecule, the analysis controller 110 can further include a clustering engine that classifies the residues in the protein molecule into one or more clusters. The molecular dynamics simulation engine 114 can perform one or more molecular dynamics simulations based on the first structural information of the first conformation 119a of the protein molecule and the first perturbation set 113a, and the evaluation engine 118 can determine, based at least on the results of the molecular dynamics simulations, whether the residues within the first perturbation set 113a can form a stable bond between the compound and the protein molecule.

[0092] FIG. 12 shows an exemplary implementation of the analysis controller 110 according to some exemplary embodiments. As shown in FIG. 12, in some cases, the molecular dynamics simulation engine 114 may include a guided docking system that can receive a first set of perturbations 113a (e.g., from the perturbation set engine 112) and structural information (e.g., an unbound coordinate file) regarding a first conformation 119a of a protein molecule in an unbound state (e.g., from the first database 1104). The guided docking system can generate a composite coordinate file that includes bound protein-structure-compound structural information, which in this case refers to a compound (e.g., a ligand) bound to the first conformation 119a of the protein molecule. For example, in some cases, the composite coordinate file can be generated by clustering individual frames of a molecular dynamics (MD) simulation based on the positions of atoms in residues included in the first set of perturbations 113a and selecting one or more clusters (e.g., the densest cluster, etc.) based on the amount of atoms included therein.

[0093] It should be understood that before generating one or more corresponding composite coordinate files via guided docking, other conformations of the protein molecule in the first database 1104 (e.g., an unbound protein structure database), such as a second conformation 119a of the same protein molecule, can be provided as input to the guided docking system. Since multiple conformations can be assumed for a single protein molecule, a single protein molecule can be associated with multiple unbound coordinate files and composite coordinate files (e.g., one composite coordinate file for each unbound coordinate file).

[0094] In some cases, the unbound coordinate file can specify the shape of a particular conformation of a protein molecule, such as the first conformation 119a, by minimally specifying the positions and orientations of all the atoms in the protein molecule. The guided docking system can generate a complex coordinate file by reducing (e.g., minimizing) the energy function associated with the compound bound to the first conformation 119a of the protein molecule, and the distances from residues within the first perturbation set 113a can be used in a scoring function (e.g., Rosetta Dock, HADDOCK, and / or Autodock, etc.). The resulting complex coordinate file can specify the positions of each atom in the bound protein-structure-compound, including, for example, the atoms within the first conformation 119a of the protein molecule and the atoms in the compound.

[0095] The molecular dynamics simulation engine 114 can include a temperature search system that can receive a complex coordinate file (e.g., including the structural information of the bound compound-protein-structure) from the guided docking system and simulate the temporal evolution of the corresponding bound protein-structure-compound. The temperature search system can receive multiple complex coordinate files simultaneously, where each complex coordinate file contains the structural information of the compound bound to a different conformation of the protein molecule. In some cases, the temperature search system can perform multiple molecular dynamics (MD) simulations at different temperatures and for different bound compound-protein-structures (e.g., complexes formed by compounds bound to different conformations of the same protein molecule) to determine the temperature at which a threshold amount of the bound compound-protein-structure dissociates.

[0096] The temporal evolution of multiple associated protein-structure-compounds (e.g., generated for different conformations of a protein molecule) can be simulated at a first temperature (e.g., 375 Kelvin). This allows obtaining multiple simulations (or trajectories) of each associated protein-structure-compound. For example, three simulations or trajectories can be simulated at the first temperature for each associated compound-protein-structure, and each simulation (or trajectory) is performed over a first predetermined period (e.g., 100 nanoseconds). If there is no trajectory that results in dissociation of the associated compound-protein-structure, one or more additional molecular dynamics simulations can be performed, and each simulation (or trajectory) is performed over a second predetermined period longer than the first predetermined period (e.g., 1 microsecond) and / or at a second temperature higher than the first temperature (e.g., 25 Kelvin higher than the first temperature).

[0097] Alternatively, if all molecular dynamics simulations (MD) or (trajectories) calculated at a given temperature result in dissociation of the associated compound-protein-structure, one or more additional molecular dynamics simulations can be performed at a second temperature lower than the first temperature (e.g., 25 Kelvin lower than the first temperature) and typically over 100 nanoseconds.

[0098] Referring again to FIG. 12, the boundary test system can determine whether the bound protein-structure-compound dissociates, for example, during one molecular dynamics simulation (or trajectory) or at the end of the simulation. The boundary test system can calculate a distance metric that indicates the distance between the first centroid (COG) of the first perturbation set 113a and the second centroid (COG) of the compound during the course of the molecular dynamics simulation. In some cases, the distance metric can correspond to the distance between the centroid (COG) of the backbone nitrogen (N) atoms within the first perturbation set 113a and the centroid (COG) of the compound. If the value of the distance metric meets one or more thresholds (e.g., exceeds the value of a threshold distance), the bound protein-structure-compound is considered to dissociate.

[0099] In some cases, a complex coordinate file can be generated for each bound compound-protein-structure where different conformations are assumed for the protein molecule. Each complex coordinate file can then be subjected to multiple molecular dynamics simulations (or trajectories) at a temperature determined to be acceptable by the temperature search system. In some cases, a new complex coordinate file can include structural information associated with the bound protein-structure-compound after (or during) the time evolution simulated by the temperature search system at a first temperature. If the corresponding bound compound-protein-structure does not dissociate at the end of the molecular dynamics simulation, the corresponding complex coordinate file can be stored in a bound structure database and further evaluated by an evaluation engine 118, for example, by calculating one or more measurement criteria (such as determining a maximum achieved score and a minimum deviation score, etc.) as a proxy for data matching between the compound and the protein molecule. Alternatively, a complex coordinate file associated with a bound compound-protein-structure that dissociates during and / or at the end of one or more molecular dynamics simulations can be discarded. Further, in some cases, an unbound coordinate file that identifies the conformation of the protein molecule when the protein molecule dissociates from the compound can be returned to an unbound protein structure database (such as a first database 1104) so that the conformation of the dissociated protein molecule can be further evaluated.

[0100] The achieved score and the deviation score can be calculated by the evaluation engine 118 and assigned to the corresponding combined compound-protein-structure complex coordinate file. The achieved score can indicate the strength of the bond between a compound and a specific conformation of a protein molecule in the corresponding compound-protein-structure. In some cases, the achieved score can be a function of (or proportional to) the number and / or packing density of hydrogen bonds of atoms within the bound compound-protein-structure at or around one or more residues in the protein molecule. The higher the value of the achieved score, the more desirable it can be as it can indicate the fit between the results of the molecular dynamics simulation and the experimental results.

[0101] The achieved score of the bound compound-protein-structure can be calculated by the following formula if the bound compound-protein-structure did not dissociate during or at the end of the molecular dynamics simulation: TIFF2025519200000015.tif6170Or, if the bound compound-protein-structure dissociated during or at the end of the molecular dynamics simulation, the achieved score of the bound compound-protein-structure can be determined based on the following equation: TIFF2025519200000016.tif5170In some exemplary embodiments, ΔE is the energy value determined for the protein molecule with and without the bound compound; Nc is the number of heavy atoms within 6.5A of the amide nitrogen atom of a residue within the perturbation set of any specific frame (or time instance) of the molecular dynamics simulation; N H is the number of nitrogen and oxygen atoms within 2.5A of the amide nitrogen atom of a residue within the perturbation set of any specific frame of the molecular dynamics simulation; d is the distance from any specific residue within the perturbation set to the geometric center of gravity (COG) of the compound (e.g., ligand) in any specific frame of the molecular dynamics simulation; <variable> is that of the molecular dynamics simulation (e.g., over the various frames of the MD simulation) It is the average of TIFF2025519200000017.tif5170.

[0102] The deviation score is given by the following formula: TIFF2025519200000018.tif6170 (where TIFF2025519200000019.tif6170 is the standard deviation of the average distance from the amide nitrogen atom (e.g., the nitrogen (N) atom contained in the amide group) within the perturbation set to the geometric center of gravity (COG) of the compound (e.g., ligand) on the orbit. More generally, during the process of molecular dynamics simulation (e.g., from one frame to another within the corresponding orbit), the distance between the amide nitrogen of the residue and the center of mass of the compound can vary. The value of the average distance can be calculated from the values of various distances (associated with different frames). Based on the average value and the values of various distances, the value of the standard deviation (or deviation score) can be calculated. The deviation score may be high during the first half of the molecular dynamics simulation rounds, but will often decrease in subsequent rounds if the compound is properly bound to the well-formed protein conformation.

[0103] In some cases, the average distance (or deviation) between the compound and the protein molecule can be determined based on the geometric center of gravity (COG) of the atoms in the compound and the geometric center of gravity (COG) of the residues within the perturbation set, or based on the average distance (or deviation) between the geometric center of the compound and each of the backbone nitrogen atoms in the perturbation set residues (described later). The geometric center of gravity (COG) of the residues within the perturbation set can be determined based on the positions of the backbone nitrogen (N) atoms in those residues. It should be understood that each residue within the perturbation set will have one backbone nitrogen (N) atom. For further explanation, TIFF2025519200000020.tif5170 shows the spatial coordinates (e.g., (x,y,z)) of each backbone nitrogen (N) atom within the perturbation set TIFF2025519200000021.tif5170 d of an arbitrary frame of simulation tt in which, the individual distances (d i,t ) can correspond to the average. TIFF2025519200000022.tifIt can be determined based on the following equation. TIFF2025519200000023.tifIn the above equation, N represents the number of residues in the perturbation set, M represents the number of frames in the trajectory, TIFF2025519200000024.tif6170or the magnitude of the variable in the units used above.

[0104] The molecular dynamics simulation engine 114 can include, for example, a decision system for determining whether to perform additional rounds of molecular dynamics simulation (MD) under different conditions. For example, if orbits of a threshold amount of different starting postures of a particular conformation of a protein molecule end in a corresponding bound compound-protein-structure that dissociates at a given temperature, one or more additional molecular dynamics simulations from the same conformation can be performed at a lower temperature (e.g., 25 K lower than the first simulation). Similarly, if orbits of a threshold amount of a threshold portion of different starting postures of a protein conformation remain associated with a particular round, the temperature can be increased to be favorable for dissociation in one or more subsequent molecular dynamics simulations.

[0105] In some cases, while the corresponding achievement score is below a threshold (e.g., less than 50, and in particular, the actual score will always be proportional to the size of the perturbation set, and this threshold will likely depend on the case) and / or the corresponding deviation score is greater than a threshold (e.g., less than 1, which is absolute and will generally not depend on the case), multiple trajectories may not end with the dissociation of the combined compound - protein - structure. In this case, one or more of the corresponding complex coordinate files can be removed to generate additional conformations of the protein molecule to be stored in a database of unbound protein structures (e.g., the first database 1104). These conformations may be different from the conformations of the protein molecule provided to the molecular dynamics simulation engine 114 for performing one or more other rounds (e.g., one or more previous rounds) of molecular dynamics simulations after a new perturbation set is generated and provided to the docking algorithm. The structural information associated with these conformations of the protein molecule can be provided as an input to the guided docking system of the molecular dynamics simulation engine 114. For example, one or more molecular dynamics simulations can be performed by the molecular dynamics simulation engine 114 for each additional conformation of the protein molecule and the corresponding perturbation set to determine whether the perturbation set can form a stable bond between the compound and the protein molecule across different conformations of the protein molecule. For example, a specific conformation of the protein molecule (e.g., the corresponding structural information such as an unbound coordinate file) and the corresponding perturbation set can be provided to the guided docking system to generate a complex coordinate file of the compound bound to the conformation of the protein molecule at the residues included in the perturbation set. When a complex coordinate file (e.g., of the conformation of the protein bound to the compound) is provided to the molecular dynamics simulation engine 114, the guided docking system can be bypassed and the complex coordinate file can be provided to the temperature search system.

[0106] The determination system can determine that a plurality of regions within the first perturbation set 113a (used by the molecular dynamics simulation engine 114 to perform simulations) are identified as forming a stable bond between the compound and the protein molecule. For example, the determination system can determine that when one or more end criteria are met, the complex coordinate file must be placed in the binding structure database and that no additional simulations are required for that molecule. In some cases, the end criteria can be based on the maximum achievement score and the minimum deviation score of one or more trajectories. For example, when the trajectory meets the achievement score criteria (e.g., the achievement score reaches a plateau state or the increase stops) and the deviation score criteria (e.g., the deviation score is less than 1), this complex coordinate file is placed in the binding structure database.

[0107] In some cases, the end criteria can also be based on the alignment between a threshold amount of structures (e.g., at least 3 or different thresholds) within the binding structure database. For example, the atoms of the protein molecule within three (or more) complex coordinate files (e.g., associated with the first trajectory and the second trajectory respectively) can be aligned, and the position of the atoms of the ligand within the first new complex coordinate file can be compared with the atoms of the ligand within the second and third new complex coordinate files. This can be achieved, for example, by taking the intersection of the sets formed by first selecting all atoms within a predetermined distance (e.g., 6 angstroms) of the ligand. The atoms within two complex coordinate files can be aligned using the minimization of the root mean square deviation (RMSD) of the carbon alpha atoms of the set intersection. From these conditions, the root mean square distance (RMSD) between the atoms in the ligand between each complex coordinate file can be calculated. If the value is less than some predetermined threshold (e.g., 1 angstrom) for a threshold amount (e.g., 3 or more) of complex coordinate files located within the binding structure database 2) If it is smaller, the termination criterion is met and the combined compound - protein - structure is placed in the result database.

[0108] In some exemplary embodiments, once the termination criterion is met, the residues within the perturbation set used for the current simulation (e.g., the first perturbation set 113a) are identified as those that form stable bonds between the compound and the protein molecule. Additionally, the perturbation set of the protein conformation and the structural information used for the current simulation (e.g., the first perturbation set 113a, the first structural information, etc.) are stored in a second database 1108 (also referred to as the selected perturbation set database).

[0109] FIG. 13 shows an exemplary implementation of a protein structure information generation algorithm 1302 that can generate multiple conformations of a protein molecule and store corresponding structure information files (e.g., unbound structure information) in a first database 1104 (e.g., an unbound protein structure database). It should be understood that the conformations of the unbound protein and the corresponding unbound structure information can be generated once (e.g., at the start of a molecular dynamics simulation) and stored in the first database 1104 (e.g., an unbound protein structure database). In some cases, the first database 1104 (e.g., an unbound protein structure database) can store multiple conformations of a protein molecule corresponding to the structural evolution of the protein molecule over time for a single protein molecule. As described above, multiple conformations of the same protein molecule structure can be received from the first database 1104 and provided as input (e.g., simultaneously) to a guided docking system. The spatial structure of the protein generating algorithm 1302 can include an accelerated molecular dynamics simulator (often referred to as accelerated molecular dynamics simulation (MD) or the metadynamics method, etc.) that can receive the unbound structure information of the protein conformation and simulate the time evolution of the initial three - dimensional structure.

[0110] From the perspective of the above-described implementation forms of the subject matter, this application discloses the following list of examples. Here, one feature of a single example, or a plurality of features of the combined examples, and optionally, a plurality of features of the examples combined with one or more features of one or more additional examples are also further examples included in the disclosure of this application.

[0111] Item 1: A computer-implemented method, comprising: (a) determining a first set of perturbations of a first conformation of a protein molecule, including identifying one or more residues to include in the first set of perturbations based at least on the protection factors of at least a portion of the residues in the protein molecule, and further excluding from the first set of perturbations at least a first residue for which the distance from the first set of perturbations to a second residue within the first set of perturbations does not correspond to one or more dimensions of a compound; (b) performing one or more molecular dynamics simulations to generate results indicating the interaction between a compound and a first plurality of residues within the first set of perturbations; and (c) identifying, based at least on the results of the one or more molecular dynamics simulations, the first plurality of residues within the set of perturbations as those forming a stable bond between the compound and the first conformation of the protein molecule.

[0112] Item 2: The method of Item 1, wherein the protection factor associated with each residue within the first conformation corresponds to the difference between a first energy of the residue in the bound state and a second energy of the residue in the unbound state.

[0113] Item 3: The method of Item 1 or 2, further comprising identifying one or more residues to include in the first set of perturbations based at least on the protection factors of one or more residues that meet at least one or more thresholds.

[0114] Item 4: The method of item 3, further comprising determining that the protection coefficient of one or more residues meets one or more thresholds based on a value of the protection coefficient that is within the percentile of the maximum protection coefficient value observed in the first conformation of the protein molecule.

[0115] Item 5: The method of item 4, further comprising determining the percentile based at least on the magnitude of the difference between the maximum protection coefficient value and a plurality of other protection coefficient values observed in the first conformation of the protein molecule.

[0116] Item 6: The method of item 4 or 5, wherein the percentile is a value between 5 and 20.

[0117] Item 7: The method according to any one of items 1 to 6, wherein the interaction between the compound and the first plurality of residues in the first perturbation set forms an induced-fit pocket within the first conformation of the protein molecule and includes at least a portion of the first plurality of residues in the first perturbation set.

[0118] Item 8: The method according to any one of items 1 to 7, wherein each run of the molecular dynamics simulation simulates the temporal evolution of the bound protein-structure-compound, which includes a compound bound to the first conformation of the protein molecule at the first plurality of residues included in the first perturbation set.

[0119] Item 9: The method of item 8, wherein the bound protein-structure-compound is generated based at least on the first perturbation set and structural information that identifies the spatial arrangement of atoms in the first conformation of the protein molecule in the unbound state.

[0120] Item 10: The method according to any one of items 1 to 9, wherein determining the first perturbation set further comprises excluding from the first perturbation set a third residue that requires a threshold amount of reorientation to interact with at least the compound.

[0121] Item 11: The method of item 10, wherein the reorientation of the threshold amount is satisfied by (i) the amount of atoms in a third residue that requires reorientation to interact with the compound, (ii) the angle of reorientation that one or more atoms in the third residue need to take to interact with the compound, and / or (iii) the distance of reorientation that one or more atoms in the third residue need to move to interact with the compound.

[0122] Item 12: The method of item 10 or 11, further comprising determining a first vector from the geometric center (COG) of a first plurality of residues in a first perturbation set to the alpha-carbon of the third residue; determining a second vector from the alpha-carbon to one of the delta-carbon, gamma-carbon, or beta-carbon present in the third residue; and determining that the reorientation of the threshold amount is satisfied based on the angle formed by the first vector and the second vector that satisfies at least one or more thresholds.

[0123] Item 13: The method according to any one of items 10 to 12, wherein the third residue is retained within the first perturbation set based at least on the third residue being glycine.

[0124] Item 14: The method according to any one of items 1 to 13, wherein determining the first perturbation set further comprises determining the distance between a first centroid of a first cluster of residues and a second centroid of a second cluster of residues, determining that the distance does not correspond to one or more dimensions of the compound, and excluding the first residue from the first perturbation set based at least on the first residue being part of the first cluster of residues.

[0125] Item 15: The method of item 14, wherein determining the first perturbation set further comprises applying a clustering technique that differentiates the first plurality of residues into at least a first cluster of residues and a second cluster of residues to the first plurality of residues.

[0126] Item 16: The method of Item 15, wherein the clustering technique includes one or more of k-means method, Mean-Shift clustering, DBSCAN (density-based spatial clustering of applications with noise), expectation maximization (EM) clustering using a Gaussian mixture model (GMM), and agglomerative hierarchical clustering.

[0127] Item 17: The method of Item 14 or 15, wherein determining the first perturbation set further includes determining a first amount of residues in a first cluster of residues and a second amount of residues in a second cluster of residues, determining a first average distance between a first centroid and the residues in the first cluster, and determining a second average distance between a second centroid and the residues in the second cluster.

[0128] Item 18: The method of Item 17, wherein determining the first perturbation set further includes excluding a first cluster of residues from the first perturbation set based on a determination that at least (i) the second cluster contains fewer residues than the first cluster, and (ii) the distance between the first centroid of the first cluster and the second centroid of the second cluster exceeds the sum of the first average distance between the first centroid and the residues in the first cluster and the second average distance between the second centroid and the residues in the second cluster.

[0129] Item 19: The method according to any one of Items 1 to 18, wherein the result includes the distance between the first geometric center of the compound and the second geometric center of a plurality of backbone nitrogen (N) atoms among the first plurality of residues in the first perturbation set.

[0130] Item 20: The method according to any one of Items 1 to 19, wherein the result includes the distance between the first geometric center of the compound and one or more amide nitrogen atoms present among the first plurality of residues in the first perturbation set.

[0131] Item 21: The method according to any one of Items 1 to 20, wherein the one or more molecular dynamics simulations include a first molecular dynamics simulation performed at a first temperature and a second molecular dynamics simulation performed at a second temperature.

[0132] Item 22: The method according to any one of Items 1 to 21, wherein the one or more molecular dynamics simulations include a first molecular dynamics simulation performed over a first length of time and a second molecular dynamics simulation performed over a second length of time.

[0133] Item 23: The method according to any one of Items 1 to 22, wherein the one or more molecular dynamics simulations include a first molecular dynamics simulation performed on a first conformation of a protein molecule associated with a first perturbation set and a second molecular dynamics simulation performed on a second conformation of the protein molecule associated with a second perturbation set.

[0134] Item 24: The method according to Item 23, wherein each of the first conformation and the second conformation includes a three-dimensional structure having a different spatial arrangement of atoms in the protein molecule.

[0135] Item 25: The method according to Item 23 or 24, further comprising determining a second perturbation set for the second conformation of the protein molecule, the determining including at least one of: (i) identifying one or more residues to include in the second perturbation set based at least on the shielding factors of at least a portion of the residues in the protein molecule; (ii) excluding from the second perturbation set a third residue whose distance from at least a fourth residue within the second perturbation set does not correspond to one or more dimensions of the compound; and (iii) excluding from the second perturbation set a fifth residue that requires a threshold amount of reorientation to interact with the compound.

[0136] Item 26: A method according to any one of Items 1 to 25, comprising a plurality of molecular dynamics simulations, wherein each simulation exposes a bound compound - protein - structure, which contains a compound bound to a protein molecule at a first plurality of residues within a first perturbation set, to different conditions.

[0137] Item 27: A method according to Item 26, wherein the result of one or more molecular dynamics simulations comprises one or more conditions under which the bound compound - protein - structure dissociates.

[0138] Item 28: A method according to Item 27, wherein the one or more conditions comprise a temperature at which the bound compound - protein - structure dissociates.

[0139] Item 29: A method according to Item 27 or 28, wherein the one or more conditions comprise a length of time before the bound compound - protein - structure dissociates.

[0140] Item 30: A method according to any one of Items 1 to 29, comprising one molecular dynamics simulation that is performed at a higher temperature than another molecular dynamics simulation in response to the bound compound - protein - structure, which contains a compound bound to a protein molecule at a first plurality of residues within a first perturbation set, not dissociating during or at the end of another molecular dynamics simulation at a threshold amount.

[0141] Item 31: A method according to any one of Items 1 to 30, comprising one molecular dynamics simulation that is performed at a lower temperature than another molecular dynamics simulation in response to the bound compound - protein - structure, which contains a compound bound to a protein molecule at a first plurality of residues within a first perturbation set, dissociating during or at the end of another molecular dynamics simulation at a threshold amount.

[0142] Item 32: The method according to any one of Items 1 to 31, including one molecular dynamics simulation that is performed over a period longer than another molecular dynamics simulation in response to a bound compound - protein - structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set not dissociating during or at the end of another molecular dynamics simulation of a threshold amount.

[0143] Item 33: The method according to any one of Items 1 to 32, including one molecular dynamics simulation that is performed over a period shorter than another molecular dynamics simulation in response to a bound compound - protein - structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set dissociating during or at the end of another molecular dynamics simulation of a threshold amount.

[0144] Item 34: The method according to any one of Items 1 to 33, wherein the first plurality of residues within the perturbation set are identified as forming a stable bond by determining one or more measurement criteria for quantifying the binding affinity between a compound and a first conformation of a protein molecule, at least based on one or more molecular dynamics simulations, and determining, at least based on the one or more measurement criteria, that the first plurality of residues within the perturbation set form a stable bond between the compound and the first conformation of the protein molecule.

[0145] Item 35: The method according to any one of Items 1 to 34, wherein determining the first perturbation set further includes determining the distance between a first residue and a second residue and excluding the first residue from the first perturbation set in response to determining that the distance between the first residue and the second residue exceeds one or more dimensions of the compound.

[0146] Item 36: A system comprising at least one data processor and at least one memory for storing instructions, wherein when the instructions are executed by the at least one data processor, the instructions perform operations including: (a) determining a first set of perturbations of a first conformation of a protein molecule, the determining of the first set of perturbations including at least identifying one or more residues to include in the first set of perturbations based on the shielding factors of at least a portion of the residues in the protein molecule, and further including excluding from the first set of perturbations at least a first residue for which the distance from the first set of perturbations to a second residue within the first set of perturbations does not correspond to one or more dimensions of the compound; (b) performing one or more molecular dynamics simulations to generate results indicative of interactions between the compound and a first plurality of residues within the first set of perturbations; and (c) identifying, based at least on the results of the one or more molecular dynamics simulations, the first plurality of residues within the set of perturbations as those forming a stable bond between the compound and the first conformation of the protein molecule.

[0147] Item 37: The method of item 36, wherein the shielding factor associated with each residue in the first conformation corresponds to the difference between a first energy of the residue in the bound state and a second energy of the residue in the unbound state.

[0148] Item 38: The system of item 36 or 37, wherein the operations further include identifying one or more residues to include in the first set of perturbations based on the shielding factors of one or more residues that meet at least one threshold.

[0149] Item 39: The system of item 38, wherein the operations further include determining that the shielding factor of one or more residues meets one or more thresholds based on the value of the shielding factor being within a percentile of the maximum shielding factor value observed in the first conformation of the protein molecule.

[0150] Item 40: The system of item 39, wherein the action further comprises determining the percentile based at least on a difference magnitude between a maximum protection coefficient value and a plurality of other protection coefficient values observed in a first conformation of a protein molecule.

[0151] Item 41: The system of item 39 or 40, wherein the percentile is a value between 5 and 20.

[0152] Item 42: The system according to any one of items 36 to 41, wherein an interaction between a compound and a first plurality of residues within a first perturbation set comprises at least a portion of the first plurality of residues within the first perturbation set that form an induced-fit pocket inside a first conformation of a protein molecule.

[0153] Item 43: The system according to any one of items 36 to 42, wherein each run of the molecular dynamics simulation simulates a temporal evolution of a bound protein-structure-compound that includes a compound bound to a first conformation of a protein molecule at a first plurality of residues included in a first perturbation set.

[0154] Item 44: The system of item 43, wherein the bound protein-structure-compound is generated based at least on a first perturbation set and structural information that identifies a spatial arrangement of atoms within a first conformation of a protein molecule in an unbound state.

[0155] Item 45: The system according to any one of items 36 to 44, wherein determining the first perturbation set further comprises excluding from the first perturbation set a third residue that requires a threshold amount of reorientation to interact with at least a compound.

[0156] Item 46: The system of Item 45, wherein the reorientation of the threshold amount is satisfied by (i) the amount of atoms in a third residue that require reorientation to interact with the compound, (ii) the angle of reorientation that one or more atoms in the third residue need to take to interact with the compound, and / or (iii) the distance of reorientation that one or more atoms in the third residue need to move to interact with the compound.

[0157] Item 47: The system of Item 45 or 46, further comprising determining a first vector from the geometric center of gravity (COG) of a first plurality of residues in a first perturbation set to the alpha-carbon of the third residue; determining a second vector from the alpha-carbon to one of the delta-carbon, gamma-carbon, or beta-carbon present in the third residue; and determining that the reorientation of the threshold amount is satisfied based on the angle formed by the first vector and the second vector that satisfies at least one or more thresholds.

[0158] Item 48: The system of any one of Items 45 to 47, wherein the third residue is retained within the first perturbation set based at least on the third residue being glycine.

[0159] Item 49: The system of any one of Items 36 to 48, wherein determining the first perturbation set further comprises determining the distance between the first centroid of the first cluster of residues and the second centroid of the second cluster of residues, determining that the distance does not correspond to one or more dimensions of the compound, and excluding the first residue from the first perturbation set based at least on the first residue being part of the first cluster of residues.

[0160] Item 50: The system of Item 49, wherein determining the first perturbation set further comprises applying a clustering technique that differentiates the first plurality of residues into at least a first cluster of residues and a second cluster of residues to the first plurality of residues.

[0161] Item 51: The system of Item 50, wherein the clustering technique includes one or more of k-means method, Mean-Shift clustering, DBSCAN (density-based spatial clustering of applications with noise), expectation maximization (EM) clustering using a Gaussian mixture model (GMM), and agglomerative hierarchical clustering.

[0162] Item 52: The system according to any one of Items 49 to 51, wherein determining the first perturbation set further includes determining a first amount of residues in a first cluster of residues and a second amount of residues in a second cluster of residues, determining a first average distance between a first centroid and the residues in the first cluster, and determining a second average distance between a second centroid and the residues in the second cluster.

[0163] Item 53: The system of Item 52, wherein determining the first perturbation set further includes excluding a first cluster of residues from the first perturbation set based on a determination that at least (i) the second cluster contains fewer residues than the first cluster, and (ii) the distance between the first centroid of the first cluster and the second centroid of the second cluster exceeds the sum of the first average distance between the first centroid and the residues in the first cluster and the second average distance between the second centroid and the residues in the second cluster.

[0164] Item 54: The system according to any one of Items 36 to 53, wherein the result includes the distance between a first geometric center of a compound and a second geometric center of a plurality of backbone nitrogen (N) atoms among a first plurality of residues in the first perturbation set.

[0165] Item 55: The system according to any one of Items 36 to 54, wherein the result includes the distance between a first geometric center of a compound and one or more amide nitrogen atoms present among a first plurality of residues in the first perturbation set.

[0166] Item 56: Any system of Items 36 to 55, wherein one or more molecular dynamics simulations include a first molecular dynamics simulation performed at a first temperature and a second molecular dynamics simulation performed at a second temperature.

[0167] Item 57: Any system of Items 36 to 56, wherein one or more molecular dynamics simulations include a first molecular dynamics simulation performed over a first length of time and a second molecular dynamics simulation performed over a second length of time.

[0168] Item 58: Any system of Items 36 to 57, wherein one or more molecular dynamics simulations include a first molecular dynamics simulation performed on a first conformation of a protein molecule associated with a first perturbation set and a second molecular dynamics simulation performed on a second conformation of the protein molecule associated with a second perturbation set.

[0169] Item 59: The system of Item 58, wherein each of the first conformation and the second conformation includes a three-dimensional structure having a different spatial arrangement of atoms in the protein molecule.

[0170] Item 60: The system of Item 58 or 59, wherein the operation further includes determining a second perturbation set of the second conformation of the protein molecule, and the determining includes at least one of: (i) identifying one or more residues to include in the second perturbation set based at least on the shielding factors of at least a portion of the residues in the protein molecule; (ii) excluding from the second perturbation set a third residue whose distance from at least a fourth residue within the second perturbation set does not correspond to one or more dimensions of the compound; and (iii) excluding from the second perturbation set a fifth residue that requires a threshold amount of reorientation to interact with at least the compound.

[0171] Item 61: A system according to any one of Items 36 to 60, comprising a plurality of molecular dynamics simulations, wherein each simulation exposes a bound compound - protein - structure, which includes a compound bound to a protein molecule at a first plurality of residues within a first perturbation set, to different conditions.

[0172] Item 62: A system according to Item 61, wherein the results of one or more molecular dynamics simulations include one or more conditions under which the bound compound - protein - structure dissociates.

[0173] Item 63: A system according to Item 62, wherein the one or more conditions include a temperature at which the bound compound - protein - structure dissociates.

[0174] Item 64: A system according to Item 62 or 63, wherein the one or more conditions include a length of time before the bound compound - protein - structure dissociates.

[0175] Item 65: A system according to any one of Items 36 to 64, comprising one molecular dynamics simulation that is performed at a higher temperature than another molecular dynamics simulation in response to the bound compound - protein - structure, which includes a compound bound to a protein molecule at a first plurality of residues within a first perturbation set, not dissociating during or at the end of another molecular dynamics simulation at a threshold amount.

[0176] Item 66: A system according to any one of Items 36 to 65, comprising one molecular dynamics simulation that is performed at a lower temperature than another molecular dynamics simulation in response to the bound compound - protein - structure, which includes a compound bound to a protein molecule at a first plurality of residues within a first perturbation set, dissociating during or at the end of another molecular dynamics simulation at a threshold amount.

[0177] Item 67: One or more molecular dynamics simulations are performed for a longer period than another molecular dynamics simulation in response to a bound compound - protein - structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set not dissociating during another molecular dynamics simulation of a threshold amount or at the end of the simulation, the system of any of items 36 to 66 comprising one molecular dynamics simulation.

[0178] Item 68: One or more molecular dynamics simulations are performed for a shorter period than another molecular dynamics simulation in response to a bound compound - protein - structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set dissociating during another molecular dynamics simulation of a threshold amount or at the end of the simulation, the system of any of items 36 to 67 comprising one molecular dynamics simulation.

[0179] Item 69: The system of any of items 36 to 68, wherein the first plurality of residues within the perturbation set are identified as forming a stable bond by determining one or more measurement criteria for quantifying the binding affinity between the compound and a first conformation of the protein molecule, at least based on one or more molecular dynamics simulations, and determining, at least based on the one or more measurement criteria, that the first plurality of residues within the perturbation set form a stable bond between the compound and the first conformation of the protein molecule.

[0180] Item 70: The system of any of items 36 to 69, wherein determining the first perturbation set further comprises determining the distance between a first residue and a second residue and excluding the first residue from the first perturbation set in response to determining that the distance between the first residue and the second residue exceeds one or more dimensions of the compound.

[0181] Item 71: A non-transitory computer-readable medium storing instructions which, when executed by at least one data processor, perform operations comprising: (a) determining a first set of perturbations of a first conformation of a protein molecule, the determining of the first set of perturbations including at least identifying one or more residues to include in the first set of perturbations based at least on shielding factors of at least a portion of the residues in the protein molecule, and further including excluding from the first set of perturbations at least a first residue for which a distance from the first set of perturbations to a second residue within the first set of perturbations does not correspond to one or more dimensions of a compound; (b) performing one or more molecular dynamics simulations to generate results indicative of interactions between a compound and a first plurality of residues within the first set of perturbations; and (c) identifying, based at least on the results of the one or more molecular dynamics simulations, the first plurality of residues within the set of perturbations as those forming a stable bond between the compound and the first conformation of the protein molecule.

[0182] Item 72: The non-transitory computer-readable medium of claim 71, wherein the instructions further perform operations including any of the methods of items 2 to 35.

[0183] FIG. 14 is a block diagram illustrating an example of a computing system 1400 according to some exemplary embodiments. In some exemplary embodiments, the computing system 1400 can be used to implement the analysis controller 110 and / or any of its components. As shown in FIG. 14, the computing system 1400 can include a processor 1410, a memory 1420, a storage device 1430, and an input / output device 1440. The processor 1410, the memory 1420, the storage device 1430, and the input / output device 1440 can be interconnected via a system bus 1450. The processor 1410 can process instructions to execute commands within the computing system. The instructions so executed can implement one or more components of, for example, the analysis controller 110 (such as the molecular dynamics simulation engine 114, the perturbation set engine 112, etc.). In some exemplary embodiments, the processor 1410 can be a single-threaded processor. Alternatively, the processor 1410 can be a multi-threaded processor. The processor 1410 can process instructions stored in the memory 1420 and / or the storage device 1430 to display graphical information for a user interface provided via the input / output device 1440.

[0184] Memory 1420 is a computer-readable medium, such as volatile or non-volatile, that stores information inside the computing system 1400. Memory 1420 can store, for example, a data structure representing a configuration object database. Storage device 1430 can provide persistent storage for the computing system 1400. The storage device 1430 can be a floppy disk device, a hard disk device, an optical disk device, a tape device, a solid state drive, and / or other suitable persistent storage means. Input / output device 1440 provides input / output operations for the computing system 1400. In some exemplary embodiments, the input / output device 1440 includes a keyboard and / or a pointing device. In various implementations, the input / output device 1440 includes a display unit for displaying a graphical user interface.

[0185] Definition Unless otherwise specified, all technical terms, notations, and other scientific or technical terms used herein are intended to have the meaning commonly understood by those of ordinary skill in the art to which this disclosure pertains. In some cases, terms with generally understood meanings are defined herein for clarity and / or ease of reference, and including such definitions herein should not necessarily be construed as representing a substantial difference from what is generally understood in the art. Many of the techniques and procedures described or referenced herein are well understood and commonly employed by those of ordinary skill in the art using conventional methodologies.

[0186] As used herein, the conformation of a protein molecule (or the conformation of a protein) is the three-dimensional structure of the residues in the sequence of amino acid residues that form the protein molecule. A residue is an organic molecule that includes an amino group, a carboxyl group, a hydrogen atom, and an alpha carbon bonded to a variable component called a side chain. As used herein, a bound protein-structure-compound is a coupling or bond between a protein molecule (having a particular conformation) and a compound (e.g., a ligand). In some cases, the protein molecule and the compound in the bound protein-structure-compound can bind to each other (e.g., the distance between the center of mass of the compound and the center of mass of a nitrogen atom within the perturbation set of the protein molecule is less than the value of a threshold distance). In some cases, the protein molecule and the compound in the bound protein-structure-compound dissociate or do not bind to each other (e.g., the distance between the center of mass of the compound and the center of mass of a nitrogen atom within the perturbation set of the protein molecule is greater than the value of a threshold distance). As used herein, the structural information associated with a particular conformation of a protein molecule includes the spatial arrangement and / or orientation of the atoms in each constituent amino acid residue that forms the protein molecule (e.g., the relative spatial positions and orientations of the atoms in the protein molecule).

[0187] The singular forms "a", "an", and "the" include plural references unless the context clearly dictates otherwise. For example, the term "a cell" includes one or more cells including mixtures thereof. As used herein, "A and / or B" is intended to include all of the following alternative forms: "A", "B", "A or B", and "A and B".

[0188] The aspects and embodiments of the disclosure described herein include aspects and embodiments "comprising", "consisting of", and / or "consisting essentially of" them.

[0189] As used herein, "comprising" is synonymous with "including", or "containing", or "characterized by", and is inclusive or non-limiting and does not exclude additional elements or method steps not recited. In particular, when the components of a composition or the steps of a method are described herein as "comprising", it is understood that compositions and methods consisting essentially of and consisting of the recited components or steps are included. As used herein, "consisting of" excludes any element, step, or component not specified in the claim for the composition or method. As used herein, "consisting essentially of" does not exclude materials or steps that do not substantially affect the basic and novel characteristics of the composition or method recited in the claim.

[0190] When a range of values is provided, those skilled in the art will understand that all ranges disclosed herein include all possible subranges and combinations of subranges. Any range recited can be readily recognized as fully describing and enabling the range to be divided at least into equal halves, thirds, quarters, fifths, tenths, etc. By way of non-limiting example, each range described herein can be readily divided into subranges such as the lower third, the middle third, and the upper third. Also, as will be understood by those skilled in the art, terms such as "maximum", "at least", "greater than", and "less than" include the recited number and refer to ranges that can be subsequently divided into subranges as described above. As will be understood by those skilled in the art, a range includes its individual members. Thus, for example, a group having 1 to 3 articles refers to a group having 1, 2, or 3 articles. Similarly, a group having 1 to 5 articles refers to a group having 1, 2, 3, 4, or 5 articles, and so on.

[0191] In some ranges, the term "about" is presented in this specification together with a numerical value that it precedes. The term "about" is used in this specification to literally support the exact number that the term precedes, as well as numbers that are close to or approximately that number which the term precedes. When determining whether a number is a number that is close to or approximately a specifically recited number, a number that is not described as close or approximate may, in the context in which it is presented, be a number that provides a substantial equivalent of the specifically recited number. When the degree of approximation is not clear from the context, "about" means within ±10% of the provided value, or rounded to the nearest significant digit in all cases that include the provided value.

[0192] Headings such as (a), (b), (i), etc. are provided only to make it easier to read this specification and the claims. The use of headings in this specification and the claims does not require that steps or elements be performed in alphabetical or numerical order, or in the order in which they are presented.

[0193] It is understood that some features of the present disclosure that are described in the context of separate embodiments for clarity may be provided in combination as a single implementation. Conversely, various features of the present disclosure that are described in the context of a single embodiment for brevity may be provided separately or in any suitable partial combination. All combinations of embodiments related to the present disclosure are specifically encompassed by the present disclosure and are disclosed herein as if each and every combination were individually and explicitly disclosed. In addition, all partial combinations of various embodiments and their elements are also specifically encompassed by the present disclosure and are disclosed herein as if each and every such partial combination were individually and explicitly disclosed herein.

[0194] A non-transitory computer program product (i.e., a physically realized computer program product) that stores instructions for causing at least one data processor to perform the operations herein when executed by one or more data processors of one or more computing systems is also described herein. Similarly, a computer system that may include one or more data processors and a memory coupled to the one or more data processors is described. The memory may store, either primarily or persistently, instructions for causing at least one processor to perform one or more of the operations described herein. Additionally, the methods may be implemented by one or more data processors, either within a single computing system or distributed among two or more computing systems. Such computing systems may be connected to exchange data and / or commands or other instructions, etc., via one or more connections including connections via a network (such as the Internet, a wireless wide area network, a local area network, a wide area network, or a wired network, etc.) such as a direct connection between one or more of the plurality of computing systems.

[0195] The subject matter described in this specification can be embodied as a system, apparatus, method, and / or article, according to a desired configuration. For example, the apparatus and / or method described herein can be implemented using one or more of processor-executable program code, application specific integrated circuits (ASICs), digital signal processors (DSPs), embedded processors, field programmable gate arrays (FPGAs), and / or combinations thereof. These various implementation forms can be executed and / or interpretable on a programmable system including at least one programmable processor that is coupled to receive data and instructions from a memory system, at least one input device, and at least one output device, and to transmit data and instructions to a memory system, at least one input device, and at least one output device, and can include an implementation into one or more computer programs. These computer programs (also known as programs, software, software applications, applications, components, program code, or code) include machine instructions for a programmable processor and can be implemented in a high-level procedural and / or object-oriented programming language and / or in assembly / machine language. As used herein, the term "machine-readable medium" refers to any computer program product, computer-readable medium, computer-readable storage medium, device, and / or device (e.g., magnetic disk, optical disk, memory, programmable logic device (PLD)) used to provide machine instructions and / or data to a programmable processor including a machine-readable medium that receives the machine instructions. Similarly, a system that can include a processor and a memory coupled to the processor is also described herein. The memory can include one or more programs that cause the processor to perform one or more of the operations described herein.

[0196] Although a few variations have been described in detail in this specification, other modifications or additions are possible. In particular, in addition to what is described in this specification, further features and / or variations may be provided. Further, the above-described implementations may also be directed to various combinations and sub-combinations of the disclosed features, and / or to multiple further combinations and sub-combinations disclosed above. In addition, the logical flows shown in the accompanying drawings and / or described in this specification do not require the particular order or sequence shown to achieve desirable results. Other implementations may be within the scope of the claims. Further, the specific values provided above are merely examples and may vary in some implementations.

[0197] While various aspects of the disclosure are shown in the claims, other aspects of the disclosure include not only the combinations explicitly shown in the claims, but also other combinations of features from the described implementations having the features of the claims.

Claims

Claim 1 A computer-implemented method comprising: (a) determining a first perturbation set of a first conformation of a protein molecule, including identifying one or more residues to include in the first perturbation set based at least on the shielding factors of at least a portion of the residues in the protein molecule, and further excluding from the first perturbation set a first residue for which the distance to at least a second residue within the first perturbation set does not correspond to one or more dimensions of a compound; (b) performing one or more molecular dynamics simulations to generate results indicative of interactions between a compound and a first plurality of residues within the first perturbation set; (c) identifying, based at least on the results of the one or more molecular dynamics simulations, the first plurality of residues within the perturbation set as forming stable bonds between the compound and the first conformation of the protein molecule; A method comprising the steps above. Claim 2 The method of claim 1, wherein the shielding factor associated with each residue in the first conformation corresponds to the difference between a first energy of the residue in the bound state and a second energy of the residue in the unbound state, and one or more residues are identified as being included in the first perturbation set based on the shielding factors of one or more residues that satisfy at least one or more thresholds. Claim 3 The method of claim 1 or 2, wherein each of the molecular dynamics simulations simulates the temporal evolution of a bound protein-structure-compound including a compound bound to the first conformation of the protein molecule at the first plurality of residues included in the first perturbation set. Claim 4 The method according to any one of claims 1 to 3, wherein determining the first perturbation set further includes excluding from the first perturbation set a third residue that requires a threshold amount of reorientation to interact with at least the compound. Claim 5 The method according to claim 4, wherein the reorientation of the threshold amount is satisfied by (i) the amount of atoms in a third residue that require reorientation to interact with the compound, (ii) the angle of reorientation that one or more atoms in the third residue need to take to interact with the compound, and / or (iii) the distance of reorientation that one or more atoms in the third residue need to move to interact with the compound.

6. Determining a first vector from the geometric center (COG) of a first plurality of residues in a first perturbation set to the alpha-carbon of a third residue; Determining a second vector from the alpha-carbon to one of the delta-carbon, gamma-carbon, or beta-carbon present in the third residue; Determining that the reorientation of the threshold amount is satisfied based on an angle formed by the first vector and the second vector that satisfies at least one or more thresholds; The method according to claim 4 or 5, further comprising:

7. Determining the first perturbation set includes: Determining the distance between a first centroid of a first cluster of residues and a second centroid of a second cluster of residues; Determining that the distance does not correspond to one or more dimensions of the compound; Excluding a first residue from the first perturbation set based on the fact that at least the first residue is part of a first cluster of residues; The method according to any one of claims 1 to 6, further comprising:

8. Determining the first perturbation set includes: Determining a first amount of residues in a first cluster of residues and a second amount of residues in a second cluster of residues; Determining a first average distance between the first centroid and the residues in the first cluster; Determining a second average distance between the second centroid and the residues in the second cluster; Excluding the first cluster of residues from the first perturbation set based on the determination that at least (i) the second cluster contains fewer residues than the first cluster and (ii) the distance between the first centroid of the first cluster and the second centroid of the second cluster exceeds the sum of the first average distance between the first centroid and the residues in the first cluster and the second average distance between the second centroid and the residues in the second cluster; The method according to claim 7, further comprising:

9. The method according to any one of claims 1 to 8, wherein the result includes at least one of (i) a first distance between a first geometric center of the compound and a second geometric center of a plurality of backbone nitrogen (N) atoms among a first plurality of residues in the first perturbation set, and (ii) a second distance between the first geometric center of the compound and one or more amide nitrogen atoms present in the first plurality of residues in the first perturbation set.

10. The method according to any one of claims 1 to 9, wherein the one or more molecular dynamics simulations include a first molecular dynamics simulation performed at a first temperature and a second molecular dynamics simulation performed at a second temperature.

11. The method according to any one of claims 1 to 10, wherein the one or more molecular dynamics simulations include a first molecular dynamics simulation performed over a first length of time and a second molecular dynamics simulation performed over a second length of time.

12. The method according to any one of claims 1 to 11, wherein the one or more molecular dynamics simulations include a first molecular dynamics simulation performed on a first conformation of a protein molecule associated with a first perturbation set and a second molecular dynamics simulation performed on a second conformation of the protein molecule associated with a second perturbation set.

13. Each of the molecular dynamics simulations exposes a bound compound-protein-structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set to different conditions, and the result of the one or more molecular dynamics simulations includes one or more conditions under which the bound compound-protein-structure dissociates, the one or more conditions including at least one of (i) a temperature at which the bound compound-protein-structure dissociates, and (ii) a length of time before the bound compound-protein-structure dissociates. The method according to any one of claims 1 to 12.

14. One or more molecular dynamics simulations are performed at a temperature higher than that of another molecular dynamics simulation in response to a bound compound - protein - structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set not dissociating during another molecular dynamics simulation of a threshold amount or at the end of the same simulation, the method according to any one of claims 1 to 13.

15. One or more molecular dynamics simulations are performed at a temperature lower than that of another molecular dynamics simulation in response to a bound compound - protein - structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set dissociating during another molecular dynamics simulation of a threshold amount or at the end of the same simulation, the method according to any one of claims 1 to 14.

16. One or more molecular dynamics simulations are performed over a period longer than that of another molecular dynamics simulation in response to a bound compound - protein - structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set not dissociating during another molecular dynamics simulation of a threshold amount or at the end of the same simulation, the method according to any one of claims 1 to 15.

17. One or more molecular dynamics simulations are performed over a period shorter than that of another molecular dynamics simulation in response to a bound compound - protein - structure comprising a compound bound to a protein molecule at a first plurality of residues within a first perturbation set dissociating during another molecular dynamics simulation of a threshold amount or at the end of the same simulation, the method according to any one of claims 1 to 16.

18. The first plurality of residues within the perturbation set comprises at least determining one or more measurement criteria for quantifying the binding affinity between a compound and a first conformation of a protein molecule, based at least on the results of one or more molecular dynamics simulations, and Determining that, based at least on one or more measurement criteria, a first plurality of residues within the perturbation set form stable bonds with a first conformation of a compound and a protein molecule The method according to any one of claims 1 to 17, identified as forming stable bonds **Claim 19** At least one data processor, At least one memory storing instructions that, when executed by at least one data processor, result in operations including the method according to any one of claims 1 to 18 A system comprising. **Claim 20** A non-transitory computer-readable medium storing instructions that, when executed by at least one data processor, result in operations including the method according to any one of claims 1 to 18

Images