CN114520031A - Method for predicting permeability of compound placental membrane based on machine learning - Google Patents
Method for predicting permeability of compound placental membrane based on machine learning Download PDFInfo
- Publication number
- CN114520031A CN114520031A CN202210079167.6A CN202210079167A CN114520031A CN 114520031 A CN114520031 A CN 114520031A CN 202210079167 A CN202210079167 A CN 202210079167A CN 114520031 A CN114520031 A CN 114520031A
- Authority
- CN
- China
- Prior art keywords
- compound
- placental membrane
- permeability
- parameter
- prediction model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 150000001875 compounds Chemical class 0.000 title claims abstract description 98
- 230000035699 permeability Effects 0.000 title claims abstract description 69
- 210000005152 placental membrane Anatomy 0.000 title claims abstract description 67
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000010801 machine learning Methods 0.000 title claims abstract description 28
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 17
- 238000012360 testing method Methods 0.000 claims abstract description 9
- 210000004369 blood Anatomy 0.000 claims abstract description 8
- 239000008280 blood Substances 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 238000005457 optimization Methods 0.000 claims description 9
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 238000011156 evaluation Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 8
- 210000002569 neuron Anatomy 0.000 claims description 7
- 239000000126 substance Substances 0.000 claims description 7
- 238000004140 cleaning Methods 0.000 claims description 5
- 238000012706 support-vector machine Methods 0.000 claims description 5
- 238000012546 transfer Methods 0.000 claims description 5
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 4
- RLFWWDJHLFCNIJ-UHFFFAOYSA-N Aminoantipyrine Natural products CN1C(C)=C(N)C(=O)N1C1=CC=CC=C1 RLFWWDJHLFCNIJ-UHFFFAOYSA-N 0.000 claims description 4
- VEQOALNAAJBPNY-UHFFFAOYSA-N antipyrine Chemical compound CN1C(C)=CC(=O)N1C1=CC=CC=C1 VEQOALNAAJBPNY-UHFFFAOYSA-N 0.000 claims description 4
- 238000001727 in vivo Methods 0.000 claims description 4
- 229960005222 phenazone Drugs 0.000 claims description 4
- 230000003169 placental effect Effects 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 210000004700 fetal blood Anatomy 0.000 claims description 3
- 230000008774 maternal effect Effects 0.000 claims description 3
- 230000007935 neutral effect Effects 0.000 claims description 3
- 150000003839 salts Chemical class 0.000 claims description 3
- 238000007477 logistic regression Methods 0.000 claims description 2
- 238000007637 random forest analysis Methods 0.000 claims description 2
- 238000005406 washing Methods 0.000 claims description 2
- 238000010606 normalization Methods 0.000 description 8
- 210000002826 placenta Anatomy 0.000 description 7
- 239000011159 matrix material Substances 0.000 description 6
- 150000002894 organic compounds Chemical class 0.000 description 5
- 238000004880 explosion Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 210000003754 fetus Anatomy 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000004888 barrier function Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 229940079593 drug Drugs 0.000 description 2
- 239000003814 drug Substances 0.000 description 2
- 238000000338 in vitro Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 230000035515 penetration Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003344 environmental pollutant Substances 0.000 description 1
- 238000010874 in vitro model Methods 0.000 description 1
- 238000005462 in vivo assay Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- -1 medicines Chemical class 0.000 description 1
- 230000001394 metastastic effect Effects 0.000 description 1
- 206010061289 metastatic neoplasm Diseases 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003121 nonmonotonic effect Effects 0.000 description 1
- 210000002220 organoid Anatomy 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 231100000719 pollutant Toxicity 0.000 description 1
- 230000035935 pregnancy Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/70—Machine learning, data mining or chemometrics
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/30—Prediction of properties of chemical compounds, compositions or mixtures
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Crystallography & Structural Chemistry (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses a method for predicting the permeability of a compound placental membrane based on machine learning, which comprises the following steps: (1) establishing a compound placental membrane permeability judgment standard; (2) acquiring a compound to establish a BPBData data set, obtaining sample data and a sample label, and preprocessing the sample data; (3) constructing a prediction model based on a machine learning algorithm, and training the prediction model under the supervision of a sample label by utilizing the preprocessed sample data so as to optimize parameters of the prediction model; (4) and predicting placental membrane permeability of the test compound. The method of the invention utilizes double parameters F/M (fetus-maternal blood concentration ratio in a compound body) and CI (clearance index) to establish a judgment standard of the permeability of the compound placental membrane; and then a prediction model of the molecular structure characteristics of the compound and the placental membrane permeability is constructed, so that the high-flux, low-time, low-cost and high-precision prediction of the compound placental membrane permeability is realized.
Description
Technical Field
The invention belongs to the technical field of compound attribute prediction, and particularly relates to a method for predicting the permeability of a compound placental membrane based on machine learning.
Background
The placenta is an important organ for intrauterine development and pregnancy maintenance, and is a structure for exchanging substances between a mother and a fetus. The placental barrier (BPB) plays an important role in the growth and development of the fetus by performing multiple functions. Since placenta is one of the most species-specific organs, in vitro models are considered to be more suitable than in vivo assays for assessing the metastatic potential of chemical substances in the human placental barrier.
However, in current studies, in vitro and ex vivo methods do not directly predict in vivo results, making assessment of the permeability of the compound to placental membranes difficult, and in vivo studies to assess the risk of transfer of chemicals from mother to fetus through placental membranes should not be performed. Under such circumstances, there is an urgent need for an evaluation method without human body test, which can achieve effective control of operability and cost while obtaining effective information on the permeability of the compound placental membrane.
Machine learning has wide application in the fields of natural language understanding, non-monotonic reasoning, machine vision, pattern recognition and the like, and can realize high-efficiency utilization of data information by deeply analyzing complex and diverse data based on the machine learning.
Chinese patent publication No. CN101339180A discloses a method for predicting the explosion characteristics of organic compounds based on a support vector machine, which includes: the parameterization of molecular structure information is realized by taking a molecular group of an organic compound as a structure descriptor for describing the molecular structure characteristics; utilizing a support vector machine to respectively simulate the internal quantitative relation between each explosion characteristic and the structure descriptor thereof, and establishing a corresponding support vector machine prediction model based on molecular groups; and inputting the molecular group of the organic compound to be predicted as an input parameter into the obtained prediction model to obtain the related explosion characteristic value. The invention can accurately and quickly predict the explosion characteristics of the organic compound according to the molecular structure of the organic compound.
Chinese patent publication No. CN113255770A discloses a method for training a compound attribute prediction model, which includes: acquiring space structure information formed by atoms and chemical bonds of a first sample compound; taking the first sample compound as an input sample and the corresponding spatial structure information as an output sample, and training to obtain a spatial structure prediction model; taking a second sample compound as an input sample and corresponding attribute information as an output sample, and training to obtain a compound attribute prediction model on the basis of the spatial structure prediction model; and predicting the attribute of the compound to be detected by using the compound attribute prediction model.
Disclosure of Invention
The invention provides a method for predicting the permeability of a compound placental membrane based on machine learning, which can realize batch prediction of the compound in a short time, is low in cost, short in time consumption and high in accuracy, overcomes the current situation that the evaluation of the permeability experiment of the current compound placental membrane is difficult, and fills up the technical blank in the aspect of building a model for predicting the permeability of the placental membrane.
The technical scheme is as follows:
a method for predicting the permeability of a compound placental membrane based on machine learning, comprising the steps of:
(1) establishing a compound placental membrane permeability judgment standard;
(2) collecting a compound to establish a BPBData data set, cleaning the data set, evaluating whether a sample in the data set has placental membrane permeability according to the standard established in the step (1), taking an evaluation result as a sample label, calculating a molecular fingerprint after deriving a SMILES expression of the sample, extracting a molecular descriptor of the sample as sample data, and preprocessing the sample data;
(3) constructing a prediction model based on a machine learning algorithm, and training the prediction model under the supervision of a sample label by utilizing the preprocessed sample data so as to optimize parameters of the prediction model;
(4) and (3) calculating the molecular fingerprint after deriving the SMILES expression of the compound to be detected, extracting the molecular descriptor as data to be detected, inputting the data to be detected into a parameter optimization prediction model, and predicting the placental membrane permeability of the compound to be detected.
In the prior art, the research on the placental membrane permeability of a compound is less, and the data acquisition is very difficult, so that the method utilizes double parameters F/M (fetal-maternal blood concentration ratio in a compound body) and CI (clearance index) to establish a compound placental membrane permeability judgment standard; classifying the samples according to whether the samples can penetrate through the placental membranes, and establishing a prediction model of the molecular structure characteristics of the compounds and the permeability of the placental membranes based on machine learning to realize batch in-vitro prediction of the permeability of the compounds of the placental membranes; the classification operation can reduce the calculation complexity and improve the calculation stability; in addition, the method does not need in-vivo experiments or establishment of organoid models, and has high prediction accuracy.
In the step (1), double parameters of F/M and CI are adopted to judge the permeability of the placental membrane of the compound, wherein F/M is the fetal-maternal blood concentration ratio in the compound body, and CI is a clearance index;
F/M-concentration of Compound in fetal blood/concentration of Compound in maternal blood
CI ═ Compound placenta penetration/antipyrine placenta penetration
Wherein the priority of the parameter F/M is greater than the priority of the parameter CI;
when the parameter F/M is less than or equal to 0.15, the compound does not have the permeability of the placental membrane, and when the parameter F/M is more than or equal to 0.3, the compound has the permeability of the placental membrane;
if the parameter F/M is not available: when the parameter CI is greater than 0.80, the compound is shown to have placental membrane permeability; when the parameter CI is less than or equal to 0.80, the compound does not have the permeability of the placental membrane.
In step (2), the compounds are collected from literature experimental data or a PubChem compound database, and the like.
Considering the generalization capability of the prediction model, various types of compounds are selected when the compounds are collected, so that the trained model has wider application range.
In the step (2), the step of cleaning the data set is as follows: filling blank values into the BPBData data set, removing inorganic substances, salts and neutral molecules, removing zero values and zero variance values and removing high correlation values so as to avoid influencing a calculation result and generating an overfitting phenomenon;
for samples with more than 1 parameter F/M: if the number of the parameters F/M is 2, taking a weighted average value; if the number of the parameters F/M is more than 2, selecting the parameter F/M with the highest occurrence frequency;
for samples with more than 1 parameter CI: if the number of the parameters CI is 2, taking a weighted average value; if the number of the parameters CI is more than 2, the parameter CI with the highest frequency of occurrence is selected.
The molecular fingerprint and the molecular descriptor can be calculated by utilizing software such as Chemopy, MoDred, RDkit and the like.
The preprocessing mode comprises normalization and normalization, wherein the normalization is to process data according to columns of a characteristic matrix and convert characteristic values of a sample into the same dimension; normalization is the processing of data according to the rows of the feature matrix, mapping the data to a specified range.
In the step (3), the preprocessed sample data is divided into a training set and a testing set, the prediction model is trained by the training set, the goodness of the prediction model is evaluated by the testing set, and the parameters of the prediction model are optimized.
In the step (3), the machine learning algorithm is selected from a random forest algorithm, a logistic regression algorithm, a naive Bayes algorithm, a support vector machine algorithm or a neural network algorithm.
Preferably, the machine learning algorithm is a neural network algorithm, and the neural network algorithm comprises an input layer, a hidden layer and an output layer; further preferably, the hidden layer is 1 layer, and the number of neuron nodes is 29.
Further preferably, the transfer function of the hidden layer is a logistic activation function, and the adam algorithm is used as a weight optimization path.
The invention also provides application of the prediction method of the compound placental membrane permeability based on machine learning in prediction of the compound placental membrane permeability.
Compared with the prior art, the invention has the beneficial effects that:
(1) the prediction method overcomes the current situation that the evaluation of the permeability experiment of the current compound placental membrane is difficult; the method can be used for predicting the placenta permeability of various compounds, such as medicines, quasi-medicines and pollutants, has wide application range, can realize batch prediction of the compounds in a short time, has low cost, short time consumption and high accuracy, and fills up the technical blank in the aspect of establishing a placental membrane permeability prediction model.
(2) The prediction accuracy of the parameter optimized prediction model can reach 0.833, the accuracy is 0.893, the recall rate is 0.847, the F1_ score is 0.870, the performance of the parameter optimized prediction model is excellent, and the prediction efficiency and accuracy are high.
(3) The invention classifies samples according to whether the compounds have the placental membrane permeability by establishing a compound placental membrane permeability judgment standard, then establishes a compound molecular structure characteristic and a prediction model of the placental membrane permeability based on a neural network, and can realize the prediction of the placental membrane transfer property only by using a SMILES expression when predicting the placental membrane permeability of the compounds.
Drawings
Fig. 1 is a flow chart of a method for predicting the permeability of a compound placental membrane based on machine learning according to the present invention.
FIG. 2 is a visualization of the classification capability of the prediction model of the present invention.
FIG. 3 is a graph of the number of hidden layer neuron nodes as a function of the accuracy of the prediction model.
Detailed Description
The invention is further elucidated with reference to the figures and the examples. It should be understood that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention.
The flow chart of the prediction method of the compound placental membrane permeability based on machine learning in the invention is shown in figure 1, and comprises the following four steps:
(1) establishing a compound placental membrane permeability judgment standard;
determining the permeability of the compound placental membrane by adopting two parameters, namely F/M (fetal-maternal blood concentration ratio in a compound body) and CI (clearance index);
F/M-concentration of Compound in fetal blood/concentration of Compound in maternal blood
Compound placenta permeability/Antipyrine (Antipyrine) placenta permeability
In the criterion of the permeability of the placental membrane, the priority of the parameter F/M is greater than the priority of the parameter CI;
when the parameter F/M is less than or equal to 0.15, the compound does not have the permeability of the placental membrane and is expressed as NC; when the parameter F/M is more than or equal to 0.3, the compound has the placental membrane permeability and is represented as C;
if the parameter F/M is not available, when the parameter CI is >0.80, the compound is indicated to have placental membrane permeability, indicated as C; when the parameter CI is less than or equal to 0.80, the compound does not have the permeability of the placental membrane and is expressed as NC.
(2) Collecting a compound to establish a BPBData data set, cleaning the data set, evaluating whether a sample in the data set has placental membrane permeability according to the standard established in the step (1), taking an evaluation result as a sample label, calculating a molecular fingerprint after deriving a SMILES expression of the sample, extracting a molecular descriptor of the sample as sample data, and preprocessing the sample data;
compounds are collected from literature experimental data or PubChem compound database, etc.
The data set cleaning method comprises the following steps: filling blank values into a BPBData data set, removing inorganic substances, salts and neutral molecules, removing zero values and zero variance values and removing high correlation values;
for samples with more than 1 parameter F/M: if the number of the parameters F/M is 2, taking a weighted average value; if the number of the parameters F/M is more than 2, selecting the parameter F/M with the highest occurrence frequency;
for samples with more than 1 parameter CI: if the number of the parameters CI is 2, taking a weighted average value; if the number of the parameters CI is more than 2, the parameter CI with the highest frequency of occurrence is selected.
After data set washing and placental membrane permeability evaluation, the BPBData data set included a total of 248 sample compounds, including 200C compounds and 48 NC compounds.
Deriving a SMILES expression and a linear molecular structure of the sample, extracting specific structural features of the sample by using RDkit software, taking the linear molecular structure of the sample as input, taking an RDkit molecular descriptor as output, wherein each column of data corresponds to one molecular descriptor, and finally obtaining 197 columns of molecular descriptors, namely a feature matrix of 248 rows and 197 columns.
And the feature matrix is preprocessed, so that the model performance loss caused by overlarge feature value difference is avoided. The preprocessing mode comprises normalization and normalization, wherein the normalization is to process data according to columns of a characteristic matrix and convert characteristic values of a sample into the same dimension; normalization is the processing of data according to the rows of the feature matrix, mapping the data to a specified range.
(3) And constructing a prediction model based on a machine learning algorithm, and training the prediction model under the supervision of a sample label by utilizing the preprocessed sample data so as to optimize parameters of the prediction model.
A prediction model is constructed by selecting a neural network, wherein the neural network comprises an input layer, a 1-layer hidden layer and an output layer.
Dividing the preprocessed sample data into a training set and a testing set, wherein the proportion of the training set to the testing set is preferably 8:2, the transfer function of the hidden layer is a logistic activation function, and adam' is selected through weight optimization; and training the prediction model by using a training set, evaluating the goodness of the prediction model by using a test set, and optimizing parameters of the prediction model.
The model goodness evaluation index comprises: accuracy, precision, recall, and F1_ score. The accuracy rate refers to the proportion of all correctly predicted sample data in the total sample data; the accuracy measures the probability that all samples predicted to be positive are true positive, and the accuracy is opposite to the false alarm rate, namely the higher the accuracy is, the less the false alarm rate is; f1_ score is a harmonic mean of accuracy and recall, and can evaluate the model more comprehensively.
In the training process, the output value of each layer is used as the node value of the next layer to continue to be calculated until the final predicted value is output, and learning and training of the model are carried out by continuously iteratively updating the weight by using an adam optimization algorithm to obtain the optimal weight.
The classification ability visualization graph of the prediction model of the present invention is shown in fig. 2, and the classification ability of the prediction model of the present invention is excellent.
And performing parameter tuning aiming at the maximum iteration number, the hidden layer neuron node number and the alpha value, wherein traversing and parameter tuning are performed on the hidden layer neuron node number. The functional relationship between the number of hidden layer neuron nodes and the accuracy of the prediction model is shown in fig. 3, and when the number of hidden layer neuron nodes is 29, the model accuracy reaches a peak value of 83.3%.
The prediction model after parameter optimization has excellent performance, the prediction accuracy of the prediction model after parameter optimization can reach 0.833, the accuracy is 0.893, the recall rate is 0.847, and the F1_ score is 0.870.
(4) And (3) calculating the molecular fingerprint after deriving the SMILES expression of the compound to be detected, extracting the molecular descriptor as data to be detected, inputting the data to be detected into a parameter optimization prediction model, and predicting the placental membrane permeability of the compound to be detected.
The embodiments described above are intended to illustrate the technical solutions of the present invention in detail, and it should be understood that the above-mentioned embodiments are only specific embodiments of the present invention, and are not intended to limit the present invention, and any modification, supplement or similar substitution made within the scope of the principles of the present invention should be included in the protection scope of the present invention.
Claims (9)
1. A method for predicting the permeability of a compound placental membrane based on machine learning, comprising the steps of:
(1) establishing a compound placental membrane permeability judgment standard;
(2) collecting a compound to establish a BPBData data set, cleaning the data set, evaluating whether a sample in the data set has placental membrane permeability according to the standard established in the step (1), taking an evaluation result as a sample label, calculating a molecular fingerprint after deriving a SMILES expression of the sample, extracting a molecular descriptor of the sample as sample data, and preprocessing the sample data;
(3) constructing a prediction model based on a machine learning algorithm, and training the prediction model under the supervision of a sample label by utilizing the preprocessed sample data so as to optimize parameters of the prediction model;
(4) and (3) calculating the molecular fingerprint after deriving the SMILES expression of the compound to be detected, extracting the molecular descriptor as data to be detected, inputting the data to be detected into a parameter optimization prediction model, and predicting the placental membrane permeability of the compound to be detected.
2. The method for predicting the placental membrane permeability of a compound based on machine learning according to claim 1, wherein in step (1), the compound placental membrane permeability is determined using two parameters, F/M and CI, wherein F/M is the fetal-maternal blood concentration ratio of the compound in vivo, and CI is the clearance index;
F/M-concentration of Compound in fetal blood/concentration of Compound in maternal blood
CI ═ Compound placental permeability/antipyrine placental permeability
Wherein the priority of the parameter F/M is greater than the priority of the parameter CI;
when the parameter F/M is less than or equal to 0.15, the compound does not have the permeability of the placental membrane, and when the parameter F/M is more than or equal to 0.3, the compound has the permeability of the placental membrane;
if the parameter F/M is not available: when the parameter CI is greater than 0.80, the compound is shown to have placental membrane permeability; when the parameter CI is less than or equal to 0.80, the compound does not have the permeability of the placental membrane.
3. The method of predicting placental membrane permeability based on machine learning of claim 1, wherein the compound is collected from literature experimental data or compound databases.
4. The method of predicting machine-learning-based compound placental membrane permeability of claim 1, wherein the step of washing the data set comprises: filling blank values into a BPBData data set, removing inorganic substances, salts and neutral molecules, removing zero values and zero variance values and removing high correlation values;
for samples with more than 1 parameter F/M: if the number of the parameters F/M is 2, taking a weighted average value; if the number of the parameters F/M is more than 2, selecting the parameter F/M with the highest occurrence frequency;
for samples with more than 1 parameter CI: if the number of the parameters CI is 2, taking a weighted average value; if the number of the parameters CI is more than 2, the parameter CI with the highest frequency of occurrence is selected.
5. The method according to claim 1, wherein in step (3), the preprocessed sample data is divided into a training set and a test set, the prediction model is trained by the training set, and the prediction model is evaluated by the test set to optimize parameters of the prediction model.
6. The method of predicting machine-learning-based compound placental membrane permeability of claim 1, wherein the machine learning algorithm is selected from a random forest algorithm, a logistic regression algorithm, a naive bayes algorithm, a support vector machine algorithm, or a neural network algorithm.
7. The method of predicting placental membrane permeability based on machine learning of claim 6, wherein the machine learning algorithm is a neural network algorithm comprising an input layer, a hidden layer and an output layer; the hidden layer is 1 layer, and the number of the neuron nodes is 29.
8. The machine learning-based compound placental membrane permeability prediction method of claim 7, wherein the transfer function of the hidden layer is a logistic activation function and adam's algorithm is used as a weight optimization path.
9. Use of a machine learning based prediction method of compound placental membrane permeability according to any one of claims 1-8 for predicting compound placental membrane permeability.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210079167.6A CN114520031B (en) | 2022-01-24 | 2022-01-24 | Prediction method of compound placenta membrane permeability based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210079167.6A CN114520031B (en) | 2022-01-24 | 2022-01-24 | Prediction method of compound placenta membrane permeability based on machine learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114520031A true CN114520031A (en) | 2022-05-20 |
| CN114520031B CN114520031B (en) | 2024-09-13 |
Family
ID=81597248
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210079167.6A Active CN114520031B (en) | 2022-01-24 | 2022-01-24 | Prediction method of compound placenta membrane permeability based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114520031B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115274002A (en) * | 2022-06-13 | 2022-11-01 | 中国科学院广州地球化学研究所 | Compound persistence screening method based on machine learning |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109214437A (en) * | 2018-08-22 | 2019-01-15 | 湖南自兴智慧医疗科技有限公司 | A kind of IVF-ET early pregnancy embryonic development forecasting system based on machine learning |
| US20210056691A1 (en) * | 2019-08-19 | 2021-02-25 | The Penn State Research Foundation | Systems and methods utilizing artificial intelligence for placental assessment and examination |
| CN112750510A (en) * | 2021-01-18 | 2021-05-04 | 合肥工业大学 | Method for predicting permeability of blood brain barrier of medicine |
-
2022
- 2022-01-24 CN CN202210079167.6A patent/CN114520031B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109214437A (en) * | 2018-08-22 | 2019-01-15 | 湖南自兴智慧医疗科技有限公司 | A kind of IVF-ET early pregnancy embryonic development forecasting system based on machine learning |
| US20210056691A1 (en) * | 2019-08-19 | 2021-02-25 | The Penn State Research Foundation | Systems and methods utilizing artificial intelligence for placental assessment and examination |
| CN112750510A (en) * | 2021-01-18 | 2021-05-04 | 合肥工业大学 | Method for predicting permeability of blood brain barrier of medicine |
Non-Patent Citations (2)
| Title |
|---|
| SIGRID CONINGS，ET AL.: "Integration and validation of the ex vivo human placenta perfusion model", 《JOURNAL OF PHARMACOLOGICAL AND TOXICOLOGICAL METHODS》, vol. 88, 15 May 2017 (2017-05-15), pages 25 - 31, XP055516507, DOI: 10.1016/j.vascn.2017.05.002 * |
| 李婷婷，等: "人胎盘的体外研究模型及其应用进展", 《中国药理学与毒理学杂志》, vol. 27, no. 6, 31 December 2013 (2013-12-31), pages 1038 - 1042 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115274002A (en) * | 2022-06-13 | 2022-11-01 | 中国科学院广州地球化学研究所 | Compound persistence screening method based on machine learning |
| CN115274002B (en) * | 2022-06-13 | 2023-05-23 | 中国科学院广州地球化学研究所 | Compound persistence screening method based on machine learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114520031B (en) | 2024-09-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Babu et al. | Medical disease prediction using grey wolf optimization and auto encoder based recurrent neural network | |
| CN110491511B (en) | Multi-model complementary enhanced machine learning method based on perioperative risk early warning | |
| CN112233736B (en) | Knowledge base construction method and system | |
| CN113679393B (en) | ECG data feature generation model based on contrast predictive coding | |
| CN110827922B (en) | Prediction method of amniotic fluid protein based on circulating neural network | |
| CN112289391A (en) | Anode aluminum foil performance prediction system based on machine learning | |
| Hossin et al. | Breast cancer detection: an effective comparison of different machine learning algorithms on the Wisconsin dataset | |
| CN114520031B (en) | Prediction method of compound placenta membrane permeability based on machine learning | |
| CN115099149A (en) | Result prediction method based on multiple feature comparison and random forest algorithm | |
| Al-Zubaidi et al. | Stroke prediction using machine learning classification methods | |
| Assegie et al. | Extraction of human understandable insight from machine learning model for diabetes prediction | |
| CN112885415A (en) | Molecular surface point cloud-based estrogen activity rapid screening method | |
| CN116702132A (en) | Network intrusion detection method and system | |
| CN115277159A (en) | Industrial Internet security situation assessment method based on improved random forest | |
| CN113936804B (en) | System for constructing model for predicting risk of continuous air leakage after lung cancer resection | |
| AV et al. | Evaluation of Recurrent Neural Network Models for Parkinson's Disease Classification Using Drawing Data | |
| CN116702839A (en) | Model training method and application system based on convolutional neural network | |
| Kadtan et al. | GUI based Prediction of Heart Stroke Stages by finding the accuracy using Machine Learning algorithm | |
| CN114360660A (en) | Method for predicting human body barrier permeability of compound based on machine learning | |
| CN113066544B (en) | FVEP characteristic point detection method based on CAA-Net and LightGBM | |
| Wójcik et al. | A complete system for an automated ECG diagnosis | |
| Tang et al. | Explainable and efficient deep early warning system for cardiac arrest prediction from electronic health records | |
| CN110942448B (en) | Quantitative phase image identification method based on convolutional neural network | |
| CN112365992A (en) | Medical examination data identification and analysis method based on NRS-LDA | |
| Bolshakova et al. | Incorporating biological domain knowledge into cluster validity assessment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |