KR102568628B1 - A method for analyzing a resistance of targeted anti-cancer therapy and identifying a combination target to overcome the resistance using network simulation - Google Patents

Abstract

A method for determining a combination target node that is combined and perturbed together with a target node that is perturbed in a cancer signaling network is disclosed. This method calculates the activity change of the candidate node based on the difference between the first activity and the second activity, which are the activities of the candidate node of the signaling network before and after the perturbation of the target node, Based on a score proportional to the activity change It is determined whether to determine the candidate node as a combination target node that perturbs together with the target node.

Description

A method for analyzing a resistance of targeted anti-cancer therapy and identifying a combination target to overcome the resistance using network simulation }

The present invention is relevant to a variety of fields including complex network science, systems biology, networks medicine, systems pharmacology and computational biology.

The present invention relates to a simulation for determining a combination of target drugs for cancer treatment using computing technology.

The present invention can be practiced using the new concepts presented by the inventor. Therefore, the new concepts described above will be described below. Not all of the material described in this chapter should be treated as prior art.

Targeted drugs aim to treat cancer by directly inhibiting oncogenes or oncogenic pathways, but unexpected drug resistance often occurs. Due to the complex dynamics of cancer signaling networks composed of complex feedback regulation, cancer cells can influence and adapt to drug therapy, leading to fundamental limitations of targeted therapies.

The effectiveness of targeted cancer therapy is limited by drug resistance, which frequently occurs in cancer cells during drug therapy. Such drug resistance is often caused by the intricate dynamics of a cancer signaling network composed of complex feedback regulatory relationships. For example, the effects of BRAF inhibitors in BRAF-mutant colorectal cancer cells are mediated by adaptive resistance resulting from rewiring of signaling networks through the following steps: 1) inhibition of BRAF and downstream ERK mitigates negative feedback from ERK to upstream epidermal growth factor receptor (EGFR), 2) increased EGFR signaling ) induces the reactivation of ERK. These negative feedback structures from downstream to upstream of the path are widespread in signaling networks. Experimental observations of anti-cancer drugs targeting some canonical signaling pathways in cancer cells show that, in general, dynamic reprogramming within the signaling network causes cancer cells to counteract the drug's effect. ), suggesting that it allows adaptation to drug perturbation. Therefore, combinatorial targeting is required to prevent such non-intuitive adaptive resistance and improve drug efficacy. Conventional screening approaches to find additional effective drug targets require a large amount of experiments, but systems analysis of signaling networks with mathematical models is a new way to understand resistance mechanisms and overcome them. It can provide a more efficient way to identify targets.

The foregoing information is provided as background technology to aid understanding of the present invention. The above contents may include contents that are known to the inventor of the present invention and have not been disclosed to unspecified persons prior to the filing date of the present patent application. Therefore, all of the above should not be considered as a matter of course that has already been disclosed.

Conventional simulation approaches have limitations in adequately capturing the dynamics of feedback regulation that causes adaptive resistance in signaling networks. Therefore, the present invention intends to provide a technique for determining combination information of a target node capable of overcoming adaptation resistance due to feedback control.

According to the present invention, it is possible to develop a network analysis of feedback regulation and identify critical combination targets to overcome adaptive resistance.

In order to identify a combinatorial drug target that can overcome the limitations due to the above problems, the present invention proposes a Boolean network simulation and analysis framework. This can be applied to large-scale signaling networks in colorectal cancer based on integrated genomic information.

In order to solve the above problems, in the present invention, the probability that the detailed mechanism of adaptive resistance to drugs can be analyzed by indicating a partially suppressed state of targeted molecules as a result of feedback regulation We propose a probabilistic Boolean simulation framework. Next, the present invention uses this framework to investigate the reconstructed large-scale signaling network of colorectal cancer, including extensive feedback regulation and crosstalks between pathways. can

<Signaling Network / Directed Code Link / Activity>

17 illustrates the concept of a biochemical reaction and directed signed link in a signal transduction network dealt with in the present invention.

In the present specification, a 'signal transduction network' refers to a network composed of interactions between various signal transduction molecules. A plurality of links may be included in one signaling network. In this case, each link may be the directed code link having directionality. Therefore, two nodes (=signaling molecules) selected in the signaling network can be connected by a directed code link. 18 schematically shows an example of a structure of a complex signaling network.

When a specific protein or compound binds to a cell membrane receptor outside the cell or when the function of a specific signaling molecule is affected by a gene mutation or drug inside the cell, a series of biochemical reactions occur, resulting in important physiological effects of the cell. function will change. A series of biochemical reactions occurring at this time may be referred to as signal transduction (signaling).

The biochemical reaction shown in FIG. 17 is simply activation of a protein by a phosphate group coupling catalyst, but when such biochemical reactions occur sequentially and change important functions of the cell, it can be interpreted in terms of 'signal transduction'. That is, it can be seen that the activated protein A 'transmits a signal' to protein B, and as shown on the right side of FIG. 17, it can be conceptually abstracted and expressed as a 'directed coded link'.

In biochemical reactions, the inactive form is also shown (e.g., protein B without a phosphate group attached), but only the active form is mainly shown in directed coded links. Numerous biochemical reactions related to signal transduction can be organized and integrated in the form of directed coded links, and as a result, a complex signal transduction network structure as shown in FIG. 18 can be obtained.

In one aspect of the present invention, the basic form of signal transduction is a biochemical reaction in which an activated signaling molecule activates another signaling molecule. For example, protein A and protein B can each be taken as examples of the above-described signaling molecules. At this time, as shown on the left side of FIG. 17, the activated protein A catalyzes a reaction in which a phosphate group binds to protein B using ATP, thereby activating protein B. Therefore, normally, protein B activity can be measured as the concentration of activated protein B (e.g., phosphorylated form of protein B). The activated protein catalyzes the activation of other proteins or enters the cell nucleus and activates the expression of genes important for cell physiology.

The directed coded link is a link representing a relationship between two nodes in the signaling network. Direction and code information may be mapped to one directed code link. The source of the first directed code link is the source node of the first directed code link, and the target of the first directed code link is the destination node of the first direct code link. That is, the direction of the directed coded link is determined from the source to the target.

As shown on the right side of FIG. 17, the code of the directed code link is determined according to the method of controlling the target by the source. There are mainly two methods in the signaling network. The first method is an activation (+) method, and the second method is an inactivation (-) method.

In relation to the directed code link, for example, a catalytic protein that catalyzes an actual biochemical reaction is referred to as the 'source' of the directed code link, and a protein to be catalyzed is referred to as the 'target' of the directed code link. can be called

In the present invention, the protein to which the catalyst is targeted can be referred to as a 'target' of the directed code link, and on the other hand, a protein (node) whose basal activity is changed by a drug administered for cancer treatment is referred to as a target node. can be called

Referring to the right side of FIG. 17, for each directed code link displayed, Node A is a source node of the directed code link and Node B is a target node of the directed code link. In this specification, the source node may be referred to as a source node, and the target node may be referred to as a target node.

The source node is a node corresponding to the starting point of the directed coded link. In the signaling network, the source node increases or decreases the activity of the target node according to the control method of the directed coded link.

The target node is a node corresponding to the arrival point of the directed coded link. In the signaling network, the activity of the target node is changed by the source node.

The signaling molecule may exist in an activated state (active state), and when activated, it is possible to perform various functions that cannot be observed before activation. For example, activated signaling molecules can activate other types of signaling molecules or promote gene expression to bring about changes in cell physiology.

Therefore, the activity, as a concept representing the 'extent' or 'quantity' of a signaling molecule in an active state, can be defined in various ways. For example, activity can be defined as the intracellular concentration (e.g., molar concentration (M)) of signaling molecules in an active state, or as an abstract variable without a specific physical unit (a variable with arbitrary units). can

According to one aspect of the present invention, when the signaling network of cells is modeled as a Boolean network, each node of the Boolean network represented by each signaling molecule has a value of 0 or 1. In this case, the activity of each node may be defined as an average of values each node had during simulation over time. Details of the activity defined in this way will be described later.

The term 'signal' is a concept that can be interpreted differently depending on the biological system to be analyzed. However, from a biological signal transduction point of view, "the concentration of signal transduction molecules in a form capable of signal transduction" can be interpreted or regarded as 'signal strength' or 'signal strength'.

According to one aspect of the present invention, when the biochemical reaction formula of signal transduction is abstracted with the directed coded link, the more abstract concept of 'activity' instead of directly corresponding the concept of signal to the concentration of signal transduction molecules. can be dealt with In other words, it can be considered that the activity (=signal) of a certain node (=source node) affects the activity (=signal) of another node (=target node) through a directed coded link. A phenomenon that affects a signal can be expressed as 'a signal flows'. In other words, it can be said that a signal flows from the source node to the target node.

According to the present invention, signals that are distinguished from each other can be mapped to each node. That is, a first signal may be mapped and defined with respect to the first node, and the first signal may be a value representing the first activity, which is the activity of the first node. Also, a second signal may be defined by being mapped to the second node, and the second signal may be a value indicating a second activity, that is, an activity of the second node.

That is, a plurality of 'signals' may be defined in the signaling network covered by the present invention. And each signal is defined differently for each node.

Therefore, in the present invention, the first signal corresponding to the first node may be defined as the activity of the first node.

As described above, activity in the signaling network refers to the degree to which a specific signaling molecule is activated to transmit a signal. The activity of a particular molecule affects the change in the activity of another molecule, which can be expressed as 'flow' in order for the signal of a particular molecule to increase or decrease the signal of another molecule.

The concept of 'a signal flows' proposed in the present invention may be interpreted as a concept that 'a change in the level of the first signal causes a change in the level of the second signal'. In addition, the above-mentioned "signal flows from the source node to the target node" means that "a change in the signal corresponding to the source node (= activity of the source node) affects the change in the signal (= activity of the target node) of the target node." It can also mean "to give".

In the present invention, 'signal flow' (signal flow) may refer to a phenomenon or influence of the activity of a certain signaling molecule on the activity of another signaling molecule.

<Development of Boolean Simulation Framework to Analyze Adaptive Resistance>

In the present invention, it is intended to find molecular targets capable of suppressing adaptive resistance to drugs in CRC. It is well known that treatment with MAPK signaling pathway inhibitors such as BRAFi or MEKi induces adaptive resistance by reactivation of MAPK signaling in CRC11, 12, and 13 . Therefore, we investigated whether BRAF or MEK inhibitors induce reactivation of their downstream molecule, ERK.

12 shows the results of monitoring the phosphorylation level of ERK for 1 hour and 48 hours by immunoblotting with the indicated antibodies.

In HCT116 cells treated with U0126 (MEKi), it was observed that pERK levels were inhibited at 1 hour (short term) and then reactivated at 48 hours (long term) (FIG. 12, top). In addition, in HT29 cells treated with vemurafenib (BRAFi), we also observed re-activation at 48 hours after pERK levels were suppressed at 1 hour treatment (Fig. 12, bottom).

13 shows an exemplary toy model of adaptive resistance in a signaling pathway with negative feedback. It can be seen that the pathway activity is initially reduced by drug perturbation but reactivated following adaptive changes.

As shown in (i) of Fig. 13, to explain this phenomenon, a signal with a negative feedback loop (NFL) composed of three cascade nodes passing signals from upstream node A to downstream nodes B and C. I made a toy model of the delivery network.

As shown in (ii) of FIG. 13, when node B in the NFL is treated with a drug, the drug target node is initially inhibited and pathway signaling is blocked.

However, as shown in (iii) of Fig. 13, inhibition of downstream node C mitigates its (downstream node C) negative feedback to upstream node A, and then activated upstream signaling It can overwhelm the drug effect on B.

As shown in Fig. 13(iv), this adaptive change in the network results in the reactivation of Node B.

As a result, the pathway can retain activity tolerant of drug treatment. This mechanism makes the NFL more resistant to drug treatment.

As shown in FIG. 14, in order to systematically investigate the effect of feedback modulation using a Boolean model, we simulated drug treatment in signaling network models with and without NFL.

In these models, an upstream node A passes a pathway signal through an intermediate node B to an output node C representing pathway activity.

As shown in FIG. 15, by deterministically simulating drug perturbation by fixing drug target node B to the OFF state, NFL has the same result as simple cascade (SC).

As shown at the bottom of FIG. 15 , there was no difference between the effects of single drug perturbation and combination drug perturbation on NFL, and this drug combination was experimentally confirmed to be effective in the MAPK pathway 13, 26 .

Area Under the Curve (AUC) for pathway inhibition demonstrated that negative feedback conditioning could not affect output after drug treatment.

These results indicate that the existing approach using deterministic drug simulation has limitations in analyzing adaptive resistance occurring in NFL and identifying synergistic combination targets.

In order to overcome the above limitation of drug simulation in the conventional approach, in the present invention, since drug target nodes may not be completely inhibited from drug treatment, different inhibition ratios for the drug target nodes are considered. By doing so, we present a Boolean simulation framework that analyzes adaptive resistance in signaling networks.

Unlike deterministic Boolean networks, where one node of the model updates its state according to that node's truth table, a probabilistic Boolean network (PBN) has one node, according to several truth tables, update its state. Allows for stochastic updates.

16 shows that the inhibition rate of the target node due to the drug treated to inhibit the activity of the target node is not conclusively fixed to a specific value, for example, a value of 1, but instead the inhibition rate is a probability value between 0 and 1 It is to explain the concept of controlling with.

As shown in Fig. 16, in the present invention, the PBN to simulate the drug target with different suppression ratios, such that the state value of the target node is stochastically suppressed according to the ratio (x) ranging from 0 to 1 during simulation. Adopt

According to one application of the present invention, if the drug target node is inhibited at a rate of 30% (x = 0.3), the state of the target node is fixed to OFF for a total of 30% of the simulation, and the simulation During the remaining 70% of , it is continuously updated according to the node's original truth table.

The present invention identifies Src as a promising combinatorial target for mitogen-activated protein kinase (MAPK) pathway inhibitors by analyzing the dynamics and structure of cell-line specific networks with integrated genomic information. . Conducting experiments based on the present invention, it can be found that combination treatment with Src inhibitors enhances the drug sensitivity of MAPK pathway inhibitors in KRAS-mutant or BRAF-mutant colon cancer cells. In addition, experiments based on the present invention can be found that Src inhibitors can increase the sensitivity of MAPK pathway inhibitors in lung and breast cancer cells with MAPK pathway alterations. In addition, experiments based on the present invention were conducted and found that Src inhibitors increased drug sensitivity of phosphatidylinositol-3 kinase (PI3K) pathway inhibitors in PI3K-mutant colon cancer cells. Based on the present invention, it can be concluded that Src is an upstream hub of feedback structures responsible for adaptive resistance emanating from those drugs targeting different signaling pathways in different cancer types.

Consequently, according to the present invention, a critical combination drug target that can overcome adaptive resistance to targeted inhibition of the mitogen-activated protein kinase (MAPK) pathway by blocking the essential feedback regulation responsible for resistance. ), Src can be found.

The systems approach according to the present invention can guide effective cancer treatment strategies with drug combinations by pinpointing the necessary feedback regulation that can be targeted to overcome adaptive resistance.

A method for determining a combined target node for a target node according to one aspect of the present invention is a method for determining a combined target node that is perturbed by being combined with a target node 12 that is perturbed in a cancer signaling network 1, wherein the A first activity, which is the activity of the candidate node 11 of the cancer signaling network 1 when the target node 12 is not perturbed, and a second activity, which is the activity of the candidate node when the target node is perturbed calculating an activity change (AC) of the candidate node based on the difference between the candidate nodes; Calculating a score proportional to the activity change; and determining whether to determine the candidate node as a combination target node that perturbs together with the target node based on the score. can include

At this time, the candidate node 11 is a node included in the negative feedback loop L1 including the candidate node 11, and when the activity of the candidate node 11 increases, the activity of the target node 12 increases. may be characterized by an increase in

At this time, by the perturbation of the target node, the activity of the target node is suppressed with a perturbation probability (x%), and the activity of the target node is suppressed with a perturbation failure probability ((100-x)%) It can be determined according to the truth table for determining the state of (provided that x is a real number from 0 to 100).

In this case, the calculating of the activity change may include setting a plurality of perturbation probabilities as the perturbation probabilities; and determining the activity change based on a value obtained by adding all the difference values calculated for each of the plurality of perturbation probabilities.

Alternatively, the calculating of the activity change may include setting the perturbation probability to 100%; and determining the activity change based on the difference value calculated with respect to the perturbation probability of 100%.

At this time, the cancer signaling network is a stochastic Boolean network, and if the activity of the target node is suppressed with a probability of x%, the target node has a value of 0, and with a probability of (100-x)%, the target node has a value of 0. When the activity of the target node is determined according to the truth table for determining the state of the cancer signaling network, the target node may have a value of 0 or 1.

At this time, the activity of the candidate node may be calculated based on a value related to the concentration of the phosphorylated form of the protein represented by the candidate node.

Alternatively, the activity of the candidate node may be calculated based on a value related to the amount of signaling molecules represented by the candidate node.

In this case, the method of determining the combination target node for the target node may include comparing the arm signaling network with a nominal network and removing incoming links of nodes where mutations occur; And a feedback length of a first negative feedback loop having the shortest feedback length among one or more negative feedback loops including the candidate node and the target node in the dark signaling network in a state in which the incoming links are removed ( Calculating LN); may be further included, and the score may be inversely proportional to the feedback length of the first negative feedback loop.

At this time, the method of determining the combination target node for the target node further comprises calculating a first bypass signaling effect (BA) from the candidate node to the apoptosis node of the cancer signaling network, It may be characterized in that the score increases as the first bypass signaling effect increases.

In this case, the step of calculating the first bypass signal transmission effect includes signal transmission excluding negative feedback loops including the target node among signal transmission pathways from the candidate node to the apoptosis node of the cancer signaling network. selecting first bypass signal transmission paths as paths; and calculating an increase in the first bypass signal transduction effect along the first bypass signal transduction pathways as the activity of the apoptosis node increases as the activity of the candidate node is inhibited. can include

At this time, the method of determining the combination target node for the target node further comprises calculating a second bypass signaling effect (BP) from the candidate node to the cell proliferation node of the cancer signaling network, It may be characterized in that the score increases as the second bypass signaling effect increases.

In this case, the step of calculating the second bypass signaling effect may include the second bypass excluding negative feedback loops including the target node among the signal transmission pathways from the candidate node to the cell proliferation node of the cancer signaling network. Selecting bypass signal transmission paths; and calculating to increase the effect of the second bypass signal transduction along the second bypass signal transduction pathways as the degree of decrease in the activity of the cell proliferation node increases as the activity of the candidate node is inhibited. can include

The framework proposed by the present invention is general and can be widely used to identify important drug targets that can overcome adaptive resistance.

According to the present invention, a combination target node corresponding to an upstream hub of feedback structures responsible for adaptive resistance resulting from such drugs targeting various signaling pathways of cancer types can be determined.

According to the present invention, by limiting the activity of the combination target node, it is possible to neutralize the adaptive resistance of drug treatment to the target node.

According to the present invention, it is intended to provide a technique for determining combination information of a target node capable of overcoming adaptation resistance due to feedback control.

1 shows a cancer signaling network 1 illustrated to explain a method for determining a combination target node provided according to an embodiment of the present invention.
FIG. 2 isolates and presents only the negative feedback loops included in the cancer signaling network shown in FIG. 1 .
3 is a flowchart illustrating a method of determining a perturbed combined target node combined with a perturbed target node in a dark signaling network according to an embodiment of the present invention.
FIG. 4 shows changes in activity of the target node and activity of candidate nodes in the first case in which the target node in FIG. 1 was not perturbed and the second case in which the target node was perturbed.
5 shows the change over time of the activity of the 14th node 14 controlled by the activity of the target node 12 when perturbation by drug 1 is applied to the target node 12 included in the first negative feedback loop. it is shown
6 is a flowchart illustrating a method for calculating a change in activity according to an embodiment of the present invention.
FIG. 7 is another representation of the two graphs shown in FIG. 4 .
8 is a flowchart illustrating a method of calculating activity change according to another embodiment of the present invention.
FIG. 9 shows a dark signaling network 1' having a state in which an incoming node of a fifteenth node 15, which is a node where a mutation has occurred, is removed among the nodes of the dark signaling network 1 shown in FIG. 1 .
FIG. 10 shows bypass signal pathways from candidate nodes of the cancer signaling network shown in FIG. 1 to apoptosis nodes and cell proliferation nodes.
11 is a flowchart illustrating a specific method of calculating a first bypass signaling effect and a second bypass signaling effect according to an embodiment of the present invention.
12 to 15 are simulation frameworks for analyzing adaptive resistance to drugs according to an embodiment of the present invention.
12 shows the results of monitoring the phosphorylation level of ERK for 1 hour and 48 hours by immunoblotting with the indicated antibodies.
13 shows an exemplary toy model of adaptive resistance in a signaling pathway with negative feedback. It can be seen that the pathway activity is initially reduced by drug perturbation but reactivated following adaptive changes.
14 shows an example of a Boolean network model of a signaling path comprising three nodes with and without negative feedback.
15 shows the results of deterministic drug simulation.
16 shows that the inhibition rate of the target node due to the drug treated to inhibit the activity of the target node is not conclusively fixed to a specific value, for example, a value of 1, but instead the inhibition rate is a probability value between 0 and 1 It is to explain the concept of controlling with.
17 illustrates the concept of a biochemical reaction and directed signed link in a signal transduction network dealt with in the present invention.
18 schematically shows an example of a structure of a complex signaling network.
19 illustrates a form of a stochastic target perturbation graph defined according to an embodiment of the present invention.
20 is a flowchart illustrating a method of calculating activity of a specific node according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, the present invention is not limited to the embodiments described herein and may be implemented in various other forms. Terms used in this specification are intended to aid understanding of the embodiments, and are not intended to limit the scope of the present invention. Also, the singular forms used herein include the plural forms unless the phrases clearly dictate the contrary.

According to an embodiment of the present invention, a method for determining a perturbed combined target node combined with a perturbed target node in a cancer signaling network may be provided.

1 shows a cancer signaling network 1 illustrated to explain a method for determining a combination target node provided according to an embodiment of the present invention.

The cancer signaling network 1 is composed of a plurality of nodes (square shape) and a plurality of links.

Different nodes can represent different genes. Alternatively, different nodes may represent protein molecules expressed by different genes.

In the example of FIG. 1 , the 31st node 31 is a cell proliferation node that determines cell proliferation, and the 33rd node 33 represents an apoptosis node that determines cell death. The higher the activity of the cell proliferation node, the greater the possibility of cell proliferation, and the higher the activity of the apoptosis node, the greater the possibility of cell death. In one embodiment, it can be seen that the lower the activity of the cell proliferation node of cancer cells and the higher the activity of the apoptosis node of the cancer cells, the higher the possibility of treating cancer due to the cancer cells.

The twelfth node 12 is a node targeted by drugs to be administered for cancer treatment, and may be referred to as a target node in the present specification. In one embodiment of the present invention, the target node is a specific node (the twelfth node 12), and a given situation may be assumed.

Cancer signaling network (1) represents the signaling network of a specific cancer cell line. At this time, the fifteenth node 15 represents a node where a mutation has occurred when compared with the normal cell signaling network, that is, the nominal signaling network.

FIG. 2 isolates and presents only the negative feedback loops included in the cancer signaling network shown in FIG. 1 .

The length of the first negative feedback loop (NFL1) shown in (a) of FIG. 2 is 4.

The length of the second negative feedback loop (NFL2) shown in (b) of FIG. 2 is 5.

The length of the third negative feedback loop (NFL3) shown in (c) of FIG. 2 is 3.

In (a) and (b) of FIG. 2, the directed coded link from the fourteenth node 14, which is the source, to the eleventh node 11, which is the target, acts in an inactivation manner to form negative feedback.

In (c) of FIG. 2, the directed coded link from the fifteenth node 15, which is the source, to the eleventh node 11, which is the target, acts in a suppressed manner to form negative feedback.

3 is a flowchart illustrating a method of determining a perturbed combined target node combined with a perturbed target node in a dark signaling network according to an embodiment of the present invention.

Hereinafter, it will be described with reference to FIGS. 1 to 3 together.

In one embodiment of the present invention, an arbitrary node constituting the cancer signaling network 1 may be selected as a candidate node for the combination target node.

However, in a preferred embodiment of the present invention, a node included in negative feedback loops including the target node 12 in the cancer signaling network 1 may be selected as the candidate node.

In addition, in a preferred embodiment of the present invention, a node included in upstream nodes traveling back along a directed code link toward the target node 12 among nodes included in the negative feedback loops is selected as the candidate node. can

In one embodiment of the present invention, the activity of the target node 12 may increase as the activity of the candidate node increases.

1 shows an example in which the candidate node is the eleventh node 11 included in the first negative feedback loop NFL1 including the candidate node 11 . Hereinafter, the eleventh node 11 may be referred to as a candidate node 11 .

In step S10 of FIG. 3, the first activity, which is the activity of the candidate node 11 of the cancer signaling network 1 when the target node 12 is not perturbed, and when the target node is perturbed An activity change (AC) of the candidate node may be calculated based on a difference between the second activities, which are the activities of the candidate node.

That is, before and after the perturbation of the target node 12 of the cancer cell network 1, the activity change (AC) of the target node 11 can be calculated.

In one embodiment, the perturbation may mean artificially applying restrictions to an equation for determining a state value of the target node 12 in a state equation for determining a state change of the dark signaling network 1 .

For example, when the cancer signaling network 1 is expressed as a Boolean network, perturbation of the target node 12 means that the state value of the target node 12 is fixed to a value of 0 with a probability of 100%, or Fixing to a value of 0 with a probability of x%, but controlling the state value of the target node 12 to be determined by the equation for determining the state value of the target node 12 with a probability of (100-x)% can mean

The activity change may be a real value.

Then, in step S20, a score proportional to the change in activity can be calculated. A specific method for calculating the score will be described later.

According to step S20, the score and the activity change (AC) may have a relationship of [Equation 1].

[Formula 1]

Score = k1 * AC (k1 is proportional constant)

The score may be a real value.

Then, in step S30, it is possible to determine whether to determine the candidate node as a combination target node that perturbs together with the target node based on the score.

For example, if the score is greater than the threshold value, the candidate node 11 may be determined to be a combination target node that perturbs along with the target node.

Alternatively, when there are multiple candidate nodes, for example, in a state in which the first and second candidate nodes are selected, the steps S10 and S20 are executed for each of the first and second candidate nodes. can As a result, the value of the first score, which is the score obtained for the first candidate node, and the second score, which is the score obtained for the second candidate node, can be compared. If the first score is greater than the second score, only the first candidate node may be determined to be the combination target node.

In one embodiment, by perturbation of the target node 12, the target node 12 is deactivated with a perturbation probability (x%), which is a configurable probability value (ie, the value of the target node is fixed to 0), The perturbation failure probability ((100-x)%) may be such that the value of the target node is determined according to a truth table for determining the state of the dark signaling network. In this case, x is a real number from 0 to 100.

FIG. 4 shows changes in activity of the target node and activity of candidate nodes in the first case in which the target node in FIG. 1 was not perturbed and the second case in which the target node was perturbed.

The horizontal axis of the graph shown in FIG. 4 represents the above-described perturbation probability from 0% to 100%, and the vertical axis represents the activity of the displayed node. Although the value of activity is presented as a value of 0 to 2, specific numerical values are presented to indicate the difference in the size of relative activity, and the range of numerical values that can have activity may be converted into other ranges and presented as needed. .

4(a) shows a case where the target node 12 is not perturbed. At this time, the value of the horizontal axis has no meaning, and the activity of the target node 12 and the activity of the candidate node 11 at this time are normalized to a reference value = 1 and presented.

(b) of FIG. 4 shows a case where a perturbation corresponding to drug 1 is applied to the target node 12. Since this perturbation lowers the activity of the target node 12, the activity of the twelfth node 12, the thirteenth node 13, or the fourteenth node 14 is basically lowered compared to FIG. . In addition, as the perturbation probability is set to a higher value, that is, as you move further to the right on the horizontal axis of the graph, the activity of the twelfth node 12, the thirteenth node 13, or the fourteenth node 14 increases as shown in FIG. Compared to (a), the drop rate increases.

However, since negative feedback exists between the fourteenth node 14 and the eleventh node (candidate node) 11, a decrease in the activity of the fourteenth node 14 causes an increase in the activity of the candidate node 11.

Therefore, when the activity of the target node 12 decreases due to the perturbation, the activity of the candidate node 11 basically increases. In addition, as described above, as the perturbation probability is set to a higher value, the drop in the activity of the fourteenth node 14 increases. Therefore, as the perturbation probability is set to a higher value, the activity of the candidate node 11 Compared to (a), the rise is increased.

Hereinafter, the effect of the above-described negative feedback over time will be described with reference to FIG. 5 .

5 shows the change over time of the activity of the 14th node 14 controlled by the activity of the target node 12 when perturbation by drug 1 is applied to the target node 12 included in the first negative feedback loop. it is shown

In (i) of FIG. 5 , perturbation is not applied to the target node 12 .

5(ii) is an initial state in which perturbation is applied to the target node 12. At this time, as shown in the lower group graph of FIG. 5, the activity of the fourteenth node 14 is lower than before the perturbation. As described in (b) of FIG. 4, this is because the activity of the 14th node 14 once decreases.

Looking at (iii) and (iv) of FIG. 5 , it can be seen that the activity of the 14th node 14 increases over time. As described in (b) of FIG. 4, since the activity of the candidate node 11 increases, this increased activity again increases the activity of the target node 12. At this time, the activity of the candidate node 11 also increases. It can be understood that this may occur when the effect of the perturbation on the target node 12 is greater than the effect of the perturbation on the target node 12.

6 is a flowchart illustrating a method for calculating a change in activity according to an embodiment of the present invention.

In one embodiment of the present invention, the step (S10) of calculating the activity change shown in FIG. 3 may include the following steps.

In step S110, a plurality of perturbation probabilities may be set as the perturbation probabilities.

In step S120, the activity change may be determined based on a value obtained by summing all the difference values calculated for each of the plurality of set perturbation probabilities.

In order to help understand the method shown in FIG. 6, it will be described with reference to FIG. 7 together.

FIG. 7 is another representation of the two graphs shown in FIG. 4 .

In (a) of FIG. 7, a first value (A1), which is a value obtained by integrating the activity of the candidate node 11 shown in (a) of FIG.

In (b) of FIG. 7, a second value (A2), which is a value obtained by integrating the activity of the candidate node 11 shown in (b) of FIG.

The activity change may be proportional to a value obtained by subtracting the first value A1 from the second value A2.

Setting a plurality of perturbation probabilities in step S110 of FIG. 6 may mean selecting some values among the probabilities 0% to 100% shown on the horizontal axis of FIG. 7, or the probabilities 0% to 100% It could mean choosing all of them. Conceptually, it is possible to select all of the probabilities of 0% to 100% to obtain the first and second values, which are the above-mentioned integral values, but in actual simulation, only a few of the probabilities of 0% to 100% may be selected. there is.

8 is a flowchart illustrating a method of calculating activity change according to another embodiment of the present invention.

In another embodiment of the present invention, the step (S10) of calculating the activity change shown in FIG. 3 may include the following steps.

In step S210, the perturbation probability may be set to 100%.

In step S220, the activity change may be determined based on the difference value calculated for the perturbation probability of 100%.

This embodiment is an example in which x = 1 million is set in the embodiment shown in FIG. 6 .

In one embodiment of the present invention, the cancer signaling network 1 may be modeled as a stochastic Boolean network. That is, the cancer signaling network 1 has only one of two numbers, 0 or 1. And at this time, the activity of the target node 12 can be suppressed with a probability of x%. That is, with a probability of x%, the value of the target node 12 is fixed to 0, and with a probability of (100-x)%, the value of the target node determines the state of the cancer signaling network 1. It is determined according to the equation of state or the truth table. That is, with a probability of (100-x)%, the target node 12 may have a value of 0 or 1.

In a preferred embodiment, the activity of the candidate node 11 may be defined as an average value of values of the candidate node 11 while performing a simulation provided according to an embodiment of the present invention described later.

Alternatively, the activity of the candidate node 11 may be calculated based on a value related to the concentration of the phosphorylated form of the protein represented by the candidate node 11 .

Alternatively, the activity of the candidate node 11 may be calculated based on a value related to the amount of signaling molecules represented by the candidate node 11 .

It can be calculated based on the value of the amount of signal transduction molecules represented by the candidate node 11 .

In addition, as shown in FIG. 3, the method for determining the combination target node may further include the following steps.

In step S40, the cancer signaling network 1 is compared with the nominal network 2, and incoming links of nodes where mutations occur may be removed.

Assuming that mutation does not occur at the fifteenth node 15 in the cancer signaling network 1 shown in FIG. 1 , the network shown in FIG. 1 can be regarded as the nominal network 2 .

FIG. 9 shows a dark signaling network 1' having a state in which an incoming node of a fifteenth node 15, which is a node where a mutation has occurred, is removed among the nodes of the dark signaling network 1 shown in FIG. 1 . By doing this, the third negative feedback loop (NFL3) shown in Fig. 2 (c) is not defined.

In a signaling network, an incoming link and/or an outgoing link may be connected to one node. An incoming link is an arrowhead or nail shape connected from one other node to the one node to represent when the activity of one other node affects the activity of the one node. can be a link An outgoing link is an arrow head shape or nail connected from one node to another node to represent when the activity of one node affects the activity of another node. It can be a shaped link.

Then, in step S50, in the dark signaling network 1' in which the incoming links are removed, one or more negative feedbacks including the candidate node 11 and the target node 12 are generated. Among the loops, the shortest feedback length (LN), which is the feedback length of the first negative feedback loop having the shortest feedback length, can be calculated.

In the example shown in FIG. 9 , one or more negative feedback loops including the candidate node 11 and the target node 12 are (a) and (b) of FIG. 2 . At this time, the first negative feedback loop having the shortest feedback length is NFL1 shown in (a) of FIG. In this case, the shortest feedback length (LN) of the first negative feedback loop having the shortest feedback length may be 4, which is the number of nodes included in the loop.

At this time, the score may be in inverse proportion to the feedback length (ex: 4) of the first negative feedback loop (NLF1). That is, the score (Score) and the shortest feedback length (LN) may have a relationship as shown in Equation 2.

[Equation 2]

Score = k2 / LN (k2 is a constant of proportionality)

The criterion according to Equation 1 and the criterion according to Equation 2 may be combined with each other. At this time, the score may have a relationship as shown in Equation 3.

[Formula 3]

Score = k3 * SC / LN (k3 is a constant of proportionality)

In addition, as shown in FIG. 3, the method for determining the combination target node may further include the following steps S60 and S70. Hereinafter, it will be described with reference to FIG. 10 together.

FIG. 10 shows bypass signal pathways from candidate nodes of the cancer signaling network shown in FIG. 1 to apoptosis nodes and cell proliferation nodes. All of the bypass signal paths may not include any negative feedback loops.

10(a) shows the first bypass signal pathways from the candidate node 11 of the cancer signaling network 1 to the apoptosis node 33. In (a) of FIG. 10, it is illustrated that there is only one first bypass signal path, but if the dark signaling network becomes more complex, a plurality of such paths may exist.

10(a) shows second bypass signal pathways from the candidate node 11 of the cancer signaling network 1 to the cell proliferation node 31. In (b) of FIG. 10, it is illustrated that there is only one second bypass signal path, but if the dark signaling network becomes more complex, there may be a plurality of such paths.

In step S60, a first bypass signaling effect (BA) from the candidate node 11 to the apoptosis node of the cancer signaling network may be calculated. The first bypass signaling effect BA may have a real value.

At this time, step S60 may include the following steps.

In step S61, among the signal transmission pathways from the candidate node 11 to the apoptosis node 33 of the cancer signaling network 1, the first bypass signal does not include a negative feedback loop therein. Delivery routes can be selected.

In step S62, for each of the first bypass signal transduction pathway, as the activity of the candidate node 11 is suppressed, the signal transduction effect is based on the degree to which the activity of the apoptosis node 33 increases. can be calculated.

For example, if the activity of the apoptosis node 33 increases as the activity of the candidate node 11 is suppressed, the signal transduction effect may have a positive value, and conversely, the activity of the apoptosis node 33 decreases. If so, the signaling effect may have a negative value.

In step S63, the first bypass signal transfer effect BA may be calculated based on the signal transfer effects obtained for each of the first bypass signal transfer paths.

For example, if the signaling effects have a positive value, the first bypass signaling effect BA will also have a positive value. Conversely, the first bypass signal transmission effect BA may have a negative value.

At this time, if the activity of the apoptosis node 33 increases as the activity of the candidate node 11 is further suppressed, the possibility of determining the candidate node 11 as a combination target node to be perturbed together with the target node 12 It will get bigger.

That is, if the first bypass signaling effect BA has a positive value, the score may be increased.

In step S70, a second bypass signaling effect (BP) from the candidate node 11 to the cell proliferation node of the cancer signaling network may be calculated. The second bypass signaling effect BP may have a real value.

At this time, step S70 may include the following steps.

In step S71, among the signal transmission pathways from the candidate node 11 to the cell proliferation node 31 of the cancer signaling network 1, the second bypass signal does not include a negative feedback loop therein. Delivery routes can be selected.

In step S72, for each of the second bypass signal transduction pathways, as the activity of the candidate node 11 is inhibited, the signal transduction effect is determined based on the degree to which the activity of the cell proliferation node 31 is inhibited. can be calculated.

For example, if the activity of the cell proliferation node 31 decreases as the activity of the candidate node 11 is suppressed, the signal transduction effect may have a positive value, and conversely, the activity of the cell proliferation node 31 increases. If so, the signaling effect may have a negative value.

In step S73, the second bypass signal transfer effect BP may be calculated based on the signal transfer effects obtained for each of the second bypass signal transfer paths.

For example, if the signaling effects have a positive value, the second bypass signaling effect BP will also have a positive value. Conversely, the second bypass signaling effect BP may have a negative value.

At this time, if the activity of the cell proliferation node 31 decreases as the activity of the candidate node 11 is further suppressed, the possibility of determining the candidate node 11 as a combination target node to be perturbed together with the target node 12 It will get bigger.

That is, if the second bypass signaling effect BP has a positive value, the score may be increased.

11 is a flowchart illustrating a specific method of calculating a first bypass signaling effect and a second bypass signaling effect according to an embodiment of the present invention.

Equation 4 can be presented by considering the above-described first bypass signaling effect BA and the second bypass signaling effect BP together.

[Formula 4]

Score = k4 * (BA + BP) (k4 is a constant of proportionality)

Equation 3 and Equation 4 can be combined and presented as Equation 5.

[Formula 5]

Score = k * SC * (BA + BP) / LN (k is a constant of proportionality)

In addition, as shown in FIG. 3, the method for determining the combination target node may further include the following step (S80).

In step S80, it is proportional to the activation change (SC), proportional to the sum of the first bypass signaling effect (BA) and the second bypass signaling effect (BP), and the shortest feedback length ( A score inversely proportional to LN) is calculated, and based on the value of the score, it is possible to determine whether to use the candidate node 11 as the combination target node.

If a plurality of candidate nodes are selected, the score may be calculated for each candidate node. At this time, the top N candidate nodes with the highest scores may be considered as the combination target nodes (N is selected from natural numbers), or all candidate nodes with scores exceeding a predetermined threshold may be considered as combination target nodes. may be

When obtaining the aforementioned minimum feedback length (LN), signaling effect (BP), and signaling effect (BA), always divide by the path length in the formula. The reason for this is, first, that there are many paths descending from an upper path to a lower level in the network, and each path is in a competitive relationship. This is because it is difficult for the activity of a substance in the upper pathway to be transferred as it is to one substance in the lower pathway. Second, research shows that short feedback is more influential. Therefore, long paths and feedbacks can be considered to have low influence, so the formulas for calculating the minimum feedback length (LN), signaling effect (BP), and signaling effect (BA) are divided by the path length. can include

In addition, when calculating the minimum feedback length (LN), only the negative feedback effect can be considered. The reason is that it has been found that the adaptive resistance that is currently being overcome can be induced in a negative feedback structure.

If the present invention is simply applied, an optimal combination target that overcomes resistance may be found by simply performing all drug combination simulations in a given network. Therefore, if the present invention is simply applied, the computational complexity increases as the network size increases, and the simulation method proposed here has limitations in that the calculation must be repeatedly performed with different degrees of suppression of drug target nodes even for the same drug combination. there is Therefore, a score capable of predicting the optimal combination target can be developed by considering the adaptive resistance mechanism.

<Example - Calculation method of activity>

The above description of the present invention is based on the premise that the calculated activity value can be used. Hereinafter, an embodiment of calculating the activity will be described.

19 illustrates a form of a stochastic target perturbation graph defined according to an embodiment of the present invention.

19 relates to specific drugs that inhibit the activation of specific proteins. The specific protein may correspond to a specific node of the Boolean bionetwork model. In addition, administering the specific drug can be understood as performing a perturbation that forcibly fixes the value of the specific node to 0.

In FIG. 19, the variable x on the horizontal axis has the following meaning.

First, the variable x has a value between 0 and 100 and has a meaning of %.

Second, the variable x indicates that the probability that the state value of the target node is always fixed to 0 due to the perturbation of the Boolean network corresponding to the administration of the specific drug is x%. At this time, with a probability of (100-x)%, the state value of the target node is not necessarily fixed to 0, but the state value of the target node is determined according to the state change equation for determining the state value of the target node. . Accordingly, the state value of the target node has either a value of 0 or 1 according to the state change equation.

The vertical axis of the graph shown in FIG. 19 represents the activity value of an arbitrary node grown when a cell's signaling network is modeled as a Boolean network. The activity may have a value of 0% to 100%, and the specific value may vary depending on the value of x. When the arbitrary node is a node that determines the fate of a cell, that is, a node that indicates cell death, proliferation, or suspension of growth, the vertical axis may be regarded as a value representing the cell fate.

20 is a flowchart illustrating a method of calculating activity of a specific node according to an embodiment of the present invention.

In step S100, a Boolean bio-signaling network model for a specific cell is determined.

The Boolean bio-signaling network model may simply be referred to as a Boolean network.

As the Boolean network is determined, a set of state change equations for determining the change over time of the node-link structure and state values of nodes of the Boolean network may be determined. Transition over time of the state value of the Boolean network can be found according to the set of state change equations. The state value of the Boolean network may change according to the flow of discrete time. One state value of the Boolean network may be presented as a combination of N binary values. Here, N may be the number of nodes included in the Boolean network.

The Boolean network may be composed of N nodes and a plurality of links connecting them. The Boolean network may have a total of 2^N initial state values. The total of 2^N initial state values may be expressed as an initial state value m (m is any one of natural numbers from 1 to 2^N).

A value of any one of the plurality of nodes is related to death, growth, or suspension of growth of the specific cell.

Each node has a value of either 0 or 1.

The specific cells may be specific cancer cells of a cancer patient or normal cells of the cancer patient.

The cancer cell Boolean network, which is the Boolean network for the specific cancer cell, may be a mutation from the normal cell Boolean network, which is the Boolean network for the normal cell. For example, a state transition equation for determining the state value of a specific node of the cancer cell Boolean network may be different from a state transition equation for determining the state value of the specific node of the normal cell Boolean network. For example, the state value of a specific node of the cancer cell Boolean network may be fixed at a specific value without changing over time.

In one embodiment, the node-link structure and connection relationship of the cancer cell Boolean network may be the same as the node-link structure and connection relationship of the normal cell Boolean network. However, the specific first node of the normal cell Boolean network may be referred to as a mutation node in the cancer cell Boolean network, and at this time, the state transition equation for determining the state value change of the specific first node and the mutation node State transition equations that determine state value changes may be different. In the cancer cell Boolean network, one or more mutations may exist.

It will be described with reference to FIG. 20 below.

The first process (P1) including a series of steps (S110), (S111), (S112), and (S113) shown in FIG. 20 may be repeatedly performed a plurality of times for different values of m. . Here, m may be any one value selected from among natural numbers from 1 to M=2^N, and N may be the total number of nodes in the Boolean network.

In step S110, the state value of the Boolean network is set to an initial state value (m, 1) (m is a value selected from among natural numbers from 1 to M=2^N).

In step S111, a computer that changes the state value of the Boolean network from the k-th state value (m, k) to the k+1-th state value (m, k+1) based on the state change equation set. Run the simulation (k = a natural number from 1 to i). Here, the first factor in ( , ) represents the initial value of the Boolean network, and the second factor represents the time. The k+1th state value (m, k+1) is a state immediately following the kth state value (m, k).

A series of i+1 state values changed from the initial state value (m, 1), including the initial state value (m, 1), can be obtained by the computer simulation.

In step S112, a plurality of (=j) state values among the i+1 state values calculated from the initial state values (m, 1) are set as a population, and the average value of the state values in the population is calculated. By calculating , the average value can be defined as 'activity (m)', which is activity with respect to the initial state value (m, 1).

Assuming that the number of nodes of the Boolean network is N, each of the (=j) state values may be an array having N elements. And the average value may be calculated individually for each of the N elements. Thus, the activity (m) can be presented as an array with N elements.

In step S113, 'activity (m, n)', which is the activity level of the nth node with respect to the initial state value (m, 1), can be determined.

The n-th element of the activity m represents the activity of the n-th node among a total of N nodes. Therefore, the activity of the n-th node with respect to the initial state value (m, 1) may be expressed as 'activity (m, n)'. In this case, n may be a natural number from 1 to N. Activities (m, n) may be scalar values.

In this case, j may be equal to i+1 or smaller than i+1. That is, the calculated plurality of (j) state values may be selected from i+1 state values calculated in step S111. In one embodiment, the plurality of (j) state values may be composed of the k th state value to the k+j−1 th state values (however, k+j−1=i+1, k> One). The value of the arbitrary node may have a value of 0 or 1. Therefore, if j = 100, the sum of the number of states (j1) in which the specific observation target node has a value of 0 and the number of states (j2) in which the specific observation target node has a value of 1 (j1 + j2 = j) may be 100 (= j). Accordingly, it can be understood that the average value may have a real value between 0 and 1. This may be expressed by replacing it with a probability value of 0% to 100%.

As described above, if the first process (P1) is repeatedly performed a plurality of times for different values of m, for example, if it is repeated M = 2^N times, the 'activity (, n)' for the nth node is M can be obtained. Here, the first factor of ( , ) of 'activity ( , n)' means the initial state value of the Boolean network, and the second factor (n) indicates a node. The average value of the M activities ( , n ) of the node n obtained in this way may be defined as the average activity of the node n, and this may be expressed as the activity n. 'Activity', which is an array having a total of N elements, can be obtained by combining the activities (n) obtained for each of the N nodes.

In one embodiment, the first process P1 may be performed under the following conditions.

First, in the computer simulation of changing from the k-th state value (m, k) to the k+1-th state value (m, k+1), the state value of a specific perturbation target node determined in advance follows the following rule (The specific perturbation target node is fixed to be the same for all k).

Second, among the set of state change equations, the state change equation for determining the state value of the specific perturbation target node is ignored, and the probability of forcibly fixing the state value of the specific perturbation target node to 0 is set to be x%.

Third, the probability that the state value of the specific perturbation target node is determined according to the state change equation for determining the state value of the specific perturbation target node among the set of state change equations is set to be (100-x)%.

That is, the probability that the specific perturbation target node is actually perturbed is set to x%.

Accordingly, for a certain k, while the state value of the Boolean network changes from the k-th state value (m, k) to the k+1-th state value (m, k+1), the specific perturbation target node The probability of being perturbed (= probability of perturbation, probability of perturbation, or perturbation rate) is fixed at x%.

In one embodiment, the first process P1 may be repeatedly performed for various different values of x. For example, it may be repeated a total of 101 times while changing x by a value of 1 from 0 to 100. Each time the first process P1 is performed once, one activity n for the node n may be calculated.

So far, an example of a method for calculating the activity of a specific node has been described with reference to FIG. 20 .

When the perturbation target node is specified, it may mean that a drug that affects the activity of the perturbation target node is specified.

The specific observation target node may be a node different from the specific perturbation target node. For example, the specific observation target node may be a node corresponding to a caspase enzyme among nodes of the Boolean network.

In addition, both the specific observation target node and the specific perturbation target node may be different from the mutation node.

The value of the specific observation target node may have a value of 0 or 1. Therefore, if j = 100, the sum of the number of states (j1) in which the specific observation target node has a value of 0 and the number of states (j2) in which the specific observation target node has a value of 1 (j1 + j2 = j) may be 100 (= j).

If the index representing the specific observation target node is defined as n, it can be understood that the above-described activity n may have a real value between 0 and 1. This may be expressed by replacing it with a probability value of 0% to 100%.

Using the above-described embodiments of the present invention, those belonging to the technical field of the present invention will be able to easily implement various changes and modifications without departing from the essential characteristics of the present invention. The content of each claim of the claims may be combined with other claims without reference relationship within the scope understandable through this specification.

Claims (13)

A method for determining a combined target node that is perturbed in combination with a perturbed target node (12) in a cancer signaling network (1) for cancer cells,
The first activity, which is the activity of the candidate node 11 of the cancer signaling network 1 when the computing device does not perturb the target node 12, and the candidate node when the target node is perturbed Calculating an activity change (AC) of the candidate node based on a difference between second activities, which are activities;
comparing, by the computing device, the dark signaling network with a nominal network, and removing incoming links of nodes where a mutation has occurred;
The computing device determines a first negative feedback loop having the shortest feedback length among one or more negative feedback loops including the candidate node and the target node in the dark signaling network in which the incoming links are removed. Calculate the feedback length (LN) of, calculate the first bypass signaling effect (BA) from the candidate node to the apoptosis node of the cancer signaling network, and calculate the cancer signaling network from the candidate node Calculating a second bypass signaling effect (BP) to the cell proliferation node; and
The computing device calculates a score based on the activity change (AC), the first bypass signaling effect (BA), and the second bypass signaling effect (BP), and based on the calculated score determining whether to determine the candidate node as a combination target node that perturbs together with the target node;
Including,
The target node, the combination target node, and the candidate node each represent a molecule of the cancer cell, and
The target node and the combination target node represent molecules whose basal activity is changed by drugs,
A method for determining a combination target node for a target node.
According to claim 1,
The candidate node 11 is a node included in a negative feedback loop L1 including the candidate node 11,
Characterized in that when the activity of the candidate node 11 increases, the activity of the target node 12 increases.
A method for determining a combination target node for a target node.
According to claim 1,
By perturbation of the target node, the activity of the target node is suppressed with a perturbation probability (x%), and the activity of the target node with a perturbation failure probability ((100-x)%) is the state of the cancer signaling network Determined according to the truth table for determining (where x is a real number from 0 to 100),
A method for determining a combination target node for a target node.
According to claim 3,
The step of calculating the activity change is,
setting, by the computing device, a plurality of perturbation probabilities as the perturbation probabilities; and
Determining, by the computing device, the activity change based on a sum of all the difference values calculated for each of the plurality of perturbation probabilities.
including,
A method for determining a combination target node for a target node.
According to claim 3,
The step of calculating the activity change is,
setting, by the computing device, the perturbation probability to 100%; and
Determining, by the computing device, the activity change based on the difference value calculated with respect to the perturbation probability of 100%
including,
A method for determining a combination target node for a target node.
According to claim 1,
The cancer signaling network is a stochastic Boolean network,
With a probability of x%, if the activity of the target node is inhibited, the target node has a value of 0,
With a probability of (100-x)%, if the activity of the target node is determined according to the truth table for determining the state of the cancer signaling network, the target node has a value of 0 or 1,
A method for determining a combination target node for a target node.
According to claim 1,
The method of determining a combination target node for a target node, wherein the activity of the candidate node is calculated based on the value of the concentration of the phosphorylation type of the protein represented by the candidate node.
According to claim 1,
The activity of the candidate node is calculated based on the value of the amount of signaling molecules represented by the candidate node, a method for determining a combination target node for a target node.
The method of claim 1, wherein the score is in inverse proportion to the feedback length of the first negative feedback loop.
The method of claim 1, wherein the score increases as the first bypass signaling effect increases.
According to claim 1,
The process of calculating the first bypass signaling effect,
The computing device selects first bypass signal transmission pathways, which are signal transmission pathways excluding negative feedback loops including the target node, among signal transmission pathways from the candidate node to the apoptosis node of the cancer signaling network. doing; and
The computing device, along the first bypass signal transduction pathways, increases the first bypass signal transduction effect as the degree of increase in the activity of the apoptosis node increases as the activity of the candidate node is suppressed. calculating;
including,
A method for determining a combination target node for a target node.
The method of claim 1, wherein the score increases as the second bypass signaling effect increases.
According to claim 1,
The process of calculating the second bypass signaling effect,
selecting, by the computing device, second bypass signal transmission pathways excluding negative feedback loops including the target node, among signal transmission pathways from the candidate node to the cell proliferation node of the cancer signaling network; and
The computing device, along the second bypass signal transduction pathways, as the activity of the candidate node is suppressed, as the degree of decrease in the activity of the cell proliferation node increases, the effect of the second bypass signal transduction is increased. calculating;
including,
A method for determining a combination target node for a target node.

Images