CN103675252A - A method for optimizing a composition composed of a plurality of chemical antiseptics on the basis of engineering modeling - Google Patents

Abstract

A method for optimizing a composition composed of a plurality of chemical antiseptics on the basis of engineering modeling comprises the following steps: step A, selecting a proper agricultural pathogen experiment system, determining at least two antiseptics; step B, individually measuring and establishing a responding curve between the pathogen inhibiting rate and the dosage for each antiseptic; step C, establishing a data driving model according to the obtained responding curves; step D, optimizing the formula so as to obtain an ideal antiseptic composition according to the data driving model. Through establishing a model of pathogen responding to a composite antiseptic, the method only needs a few experiments to find an optimum antiseptic composition, improves the efficiency on inhibiting pathogen, reduces the antiseptic using amount compared to that of application of single antiseptic with same efficiency, thus reduces the harms caused by chemical antiseptics to the environment, furthermore, the method adopts a plurality of antiseptics, so that the drug resistance is greatly reduced after chemical antiseptics are used for a while.

Description

The optimization method of the number of chemical germifuge combination based on Engineering Modeling

Technical field

The present invention relates to combinatorial chemistry germifuge, particularly a kind of optimization method of the number of chemical germifuge combination based on Engineering Modeling.

Background technology

In agricultural, specializes in chemistry germifuge is to the inhibition of pathogen or while killing problem, and related experiment result shows that number of chemical germifuge is combined in the effect of inhibition single germifuge relative to resistance to the action of a drug aspect and has obvious advantage.Yet along with the increase of combined bactericide quantity and every kind of germifuge dosage number, all possible number of combinations of various sterilization agent also can be exponent increase.Like this, using traditional means of experiment to screen one by one and finally obtain optimum combination is a quite thing for difficulty.

The way addressing this problem in agricultural is at present mainly to use a small amount of germifuge (being generally two kinds) to combine, every kind of germifuge is chosen a small amount of (2-3) concentration value simultaneously, thereby reduced total germifuge number of combinations, then method by experiment filters out good germifuge combination one by one.Use such method, although can find out, compare the better germifuge combination of single germifuge effect, have two defects: the one, to spend more human and material resources and time comparatively speaking; The 2nd, the concentration of sterilant screening so comparatively speaking degree of accuracy is lower, and can not guarantee it is best combination.

Along with the development of social productive forces, the disease that needs how more effective germifuge to prevent and treat crops, particularly in order to delay the drug-fast generation of pathogen, it is good solution route that many chemical bactericides are combined use.Traditional screening technique has obvious deficiency in the optimization problem of extensive many germifuge combinations.

Summary of the invention

Object of the present invention, for a kind of fast Optimization of the number of chemical germifuge combination based on Engineering Modeling is provided, the combined bactericide test of realization based on a small amount of just can optimize effective exactly, and consumption lacks and the low concentration combination of cost.

To achieve these goals, the present invention has adopted following technical scheme: a kind of optimization method of the number of chemical germifuge combination based on Engineering Modeling, comprises the steps:

A, select suitable agriculture pathogen test system, and determine to this pathogen effectively and be suitable for the germifuge that combines;

B, respectively measure every kind of germifuge to the inhibiting rate of pathogen the response curve about dosage;

C, according to the dosage range of every kind of germifuge, select suitable concentration to combine, and select a small amount of combination to carry out sterilization experiment, according to the result of a small amount of combination sterilization experiment, set up data-driven model, the response of pathogen to combined bactericide described;

D, the data-driven model based on setting up, by computer simulation, the germifuge combination that optimization is desirable.

The specific practice of steps A is for the pathogen that will prevent and treat, first to determine the sample of experiment and the environment of experiment, and approach with large Tanaka's pathogen sample and environment as far as possible; Next is to determine that determined germifuge chemical reaction can not occur after combination and causes producing new chemical composition to the known at least two kinds of germifuge that have clearly inhibition or killing effect of this pathogen.

The concrete practice of step B is, for the germifuge of every kind of selected combination, selects multiple different mass action in the pathogen that will suppress, and obtains this germifuge to the resultful concentration range of this pathogen, comprises Cmin and Cmax; And based on experimental data matching obtain every kind of germifuge to the inhibiting rate of pathogen the response curve about dosage.

The concrete practice of step C is:

C1, based on every kind of germifuge to the inhibiting rate of pathogen the response curve about dosage, evenly select several dosage, the effect of these dosage can cover inhibiting rate from 0 to possible experiment maximal value;

C2, the dosage of selected every kind of germifuge is carried out to combination in any, obtain all possible concentration combination, from these combinations, with random some combinations of selecting of equally distributed pattern, test, and measure these and combine corresponding inhibiting rate;

C3, utilize the algorithm of machine learning to set up the model between combined bactericide concentration and inhibiting rate, use the parameter in these data-optimized models that obtained, make predicted value and the experiment value of this model have best correlativity, thereby set up the data-driven model of this pathogen to combined bactericide response.

The concrete practice of step D is:

D1, by data-driven model, dope inhibiting rate corresponding to all combinations, and from wherein finding out inhibiting rate and germifuge dosage, all meet the germifuge combination of intended target;

D2, for the germifuge of finding out, combine corresponding set and carry out cluster analysis, draw the subset that in each class, germifuge constitutes, the concentration of determining every kind of germifuge in each subset is interval, and in concentration of sterilant interval corresponding to all subsets, filter out meet expectation value most part as best germifuge combination.

The present invention sets up the data-driven model of pathogen to combined bactericide response by the algorithm of machine learning, only need the experiment number of seldom measuring (for total concentration of sterilant number of combinations) just can find best germifuge combination, not only can realize high suppression efficiency, and all decline compared to the dosage that the single germifuge of same efficiency is made every kind of germifuge of used time.In addition, method of the present invention is a general method to a certain extent, similarly in problem, can use for reference use.

Accompanying drawing explanation

Fig. 1 is that in one embodiment of the invention, selected germifuge Amici reaches the response curve about dosage to the inhibiting rate of corn southern leaf blight;

Fig. 2 be in one embodiment of the invention selected germifuge polyoxin to the inhibiting rate of corn southern leaf blight the response curve about dosage;

Fig. 3 be in one embodiment of the invention selected germifuge like seedling to the inhibiting rate of corn southern leaf blight the response curve about dosage.

Embodiment

Below in conjunction with specific embodiment, the present invention is described in detail.

Embodiment 1

1, experimentation

This tests to find the optimum combination concentration of inhibition corn southern leaf blight (South Corn Leaf Blight) is target.Use southern corn leaf blight (O microspecies) as the bacterial strain of this experiment, this bacterial strain is stored in Shanghai Communications University's agricultural and biological institute.The inoculation taking a morsel is placed in the constant temperature camera bellows of 28 ℃ and cultivates after 7 days on potato glucose agar medium, uses sterilized water to collect spore and is also made into the spore suspension that concentration is every milliliter of 1 * 106 spore.On the flat board that the spore suspension of at every turn getting 50 μ L is 9cm at diameter, cultivate 5 days, to guarantee that the activity of the bacterium that each experiment is used is identical.For the experiment of each combined bactericide, the bacterial plaque that is all 7mm by diameter inoculation is dull and stereotyped upper, wherein comprises the nutrient culture media of 20mL and the combined bactericide of 200 μ L, is then placed in interior the cultivation 5 days of constant temperature camera bellows of 28 ℃.

Use right-angled intersection method to measure the diameter of bacterial plaque, use following formula to calculate corresponding inhibiting rate (IR).

$\mathrm{IR}=\frac{D_{\mathrm{control}}-D}{D_{\mathrm{control}}-D_{\mathrm{original}}}\×100\%$

Wherein D and D _controlrespectively the diameter of experimental group and control group bacterial plaque, D _originalequal 7mm.

2, the germifuge using in experiment

Totally three kinds, the germifuge using in experiment: Amici reaches, polyoxin, love seedling.They can suppress southern corn leaf blight, but the mechanism suppressing is different.It is by suppressing mitochondrial respiration that Amici reaches, thereby stops the synthetic of atriphos (ATP), and then kill fungi.It is chitinous synthetic that polyoxin can be upset cell membrane.Like that seedling is syntheticly to carry out Antifungi by what suppress ergosterol in cell membrane.

3, the optimal combination of three kinds of germifuge

1) measure respectively three kinds of germifuge to the inhibiting rate of fungi the response curve (as shown in Figure 1, Figure 2, Figure 3 shows) about dosage

2) by 1) in curve, to every kind of some concentration of germifuge uniform design, wherein Amici reaches with polyoxin and selects respectively 9 concentration values, and likes that seedling selects 10 concentration values, forms altogether 810 concentration combination.

3) in 810 combinations, 130 of random selections are tested, and show that each combines corresponding inhibiting rate.Use the model of these data training based on support vector machine (SVM), draw optimum model parameter.From remaining combination, select again 20 somes prediction inhibiting rate accuracy of model of having carried out experimental verification.

4) use a model and dope inhibiting rate corresponding to all concentration combination, and therefrom find out inhibiting rate more than 90% simultaneously the using dosage of every kind of germifuge be less than its combination (totally 137) of peaked 1/3 in experiment.

5) use k-means clustering algorithm, 137 combinations are divided into 3 classes, and draw respectively between its corresponding germifuge combination region, as shown in table 1.

Table 1

From table 1, value corresponding to C class is the expectation that more meets us, therefore elect the final result of experiment as.

Table 2 is to combine through optimizing three kinds of chemical bactericides that obtain inhibiting rate and the independent comparison (using identical concentration) of using the inhibiting rate providing of three kinds of chemical bactericides providing.

Dosage unit in table is μ gmL ^-1, inhibiting rate unit is %.

Table 2

。

Claims (5)

1. an optimization method for the combination of the number of chemical germifuge based on Engineering Modeling, is characterized in that, comprises the steps:

A, select suitable agriculture pathogen test system, and determine to this pathogen effectively and be suitable for the germifuge that combines;

B, respectively measure every kind of germifuge to the inhibiting rate of pathogen the response curve about dosage, determine the dosage range of every kind of germifuge;

C, according to the dosage range of every kind of germifuge, select suitable concentration to combine, and select a small amount of combination to carry out sterilization experiment, according to the result of a small amount of combination sterilization experiment, set up data-driven model, the response of pathogen to combined bactericide described;

D, the data-driven model based on setting up, by computer simulation, the germifuge combination that optimization is desirable.

2. method according to claim 1, is characterized in that, the specific practice of steps A is for the pathogen that will prevent and treat, first to determine the sample of experiment and the environment of experiment, and approach with large Tanaka's pathogen sample and environment as far as possible; Next is to determine that determined germifuge chemical reaction can not occur after combination and causes producing new chemical composition to the known at least two kinds of germifuge that have clearly inhibition or killing effect of this pathogen.

3. method according to claim 1, it is characterized in that, the concrete practice of step B is, germifuge for every kind of selected combination, select multiple different mass action in the pathogen that will suppress, obtain this germifuge to the resultful concentration range of this pathogen, comprise Cmin and Cmax; And based on experimental data matching obtain every kind of germifuge to the inhibiting rate of pathogen the response curve about dosage.

4. method according to claim 1, is characterized in that, the concrete practice of step C is:

C1, based on every kind of germifuge to the inhibiting rate of pathogen the response curve about dosage, evenly select several dosage, the inhibition of these dosage can cover inhibiting rate from 0 to possible experiment maximal value;

C2, the dosage of selected every kind of germifuge is carried out to combination in any, obtain all possible concentration combination, from these combinations, with the random a small amount of combination of selecting of equally distributed pattern, test, and measure these and combine corresponding inhibiting rate;

C3, utilize the algorithm of machine learning to set up the model between combined bactericide concentration and inhibiting rate, use the parameter in these data-optimized models that obtained, make predicted value and the experiment value of this model have best correlativity, thereby set up the data-driven model of this pathogen to combined bactericide response.

5. method according to claim 1, is characterized in that, the concrete practice of step D is:

D1, by data-driven model, dope inhibiting rate corresponding to all combinations, and from wherein finding out inhibiting rate and germifuge dosage, all meet the germifuge combination of intended target;

D2, for the germifuge of finding out, combine corresponding set and carry out cluster analysis, draw the subset that in each class, germifuge constitutes, the concentration of determining every kind of germifuge in each subset is interval, and in concentration of sterilant interval corresponding to all subsets, filter out meet expectation value most part as best germifuge combination.

Images