SlideShare a Scribd company logo
1 of 48
Download to read offline
1 / 47
NOVELALGORITHMS FOR DETECTION OF
UNKNOWN CHEMICAL MOLECULES WITH
SPECIFIC BIOLOGICAL ACTIVITIES
By
Ahmed Hassan Ahmed
Supervisors
Prof. Dr. Effat A. Saied Prof. Dr. Aboul Ella Hassanien
2 / 47
Agenda
• Introduction
– Chemo-informatics
– Drug discovery and its cyclic
– Molecular representation
• Problem Statement
• The proposed approaches
• Qualitative structure-activity relationships approach
• Graph algorithms based-approach
• Coding System
• A new atoms similarity algorithm
• A new paths of stars algorithm
• Conclusion and future work
3 / 47
Introduction
Chemo-informatics
Chemistry
Problems
Information
Problems
Cheminformatics is the use of
computer and informational
techniques applied to a range of
problems in the field of
chemistry.
Chemoinformatics / cheminformatics
can be broadly defined as the field of
solving chemical problems with
computers.
4 / 47
Introduction
Drug Discovery and its cycle
• Drug discovery is the process through which potential new
medicines are identified.
https://en.wikipedia.org/wiki/Drug_discovery
5 / 47
Introduction
Molecular Activity and Similarity
• Similarity principle:
• Structurally similar
molecules tend to have
similar properties
• Properties: biological activity,
solubility, color and so on.
6 / 47
Introduction
Molecular Representation
• The structure of a molecule can
be represented by a graph
• Graph = collection of nodes and
edges, nodes and edges have
properties (atomic number,
bond order).
• The molecular graph can be
presented by Connection table
which describes the relationships
between nodes.
7 / 47
Problem Statement
• The development of a new
chemical drug is still a
challenging, time-consuming
and costly process.
• This thesis aims to provide
new approaches to reduce the
time and the cost of chemical
compounds classification.
Thousands of
Candidate
Compounds
Hundreds of
Candidate
Compounds
Approved Drug
8 / 47
The Proposed Approaches
• Molecular Descriptors and Quantitative-structure
activity relationships (QSAR-Model)
• Predicting biological activity of 2,4,6-trisubstituted 1,3,5-triazines using random
forest, in: Proceedings of the 5th International Conference on Innovations in Bio-
Inspired Computing and Applications (IBICA 2014), 2014, pp. 101–110.
• Graph Algorithms
• A new atoms similarity algorithm
• Predicting activity approach based on new atoms similarity kernel function. Journal of
Molecular Graphics and Modelling, Volume 60, July 2015, Pages 55 – 62. (IF 1.754)
• A new paths of stars algorithm
• Two-class support vector machine with new kernel function based on paths of features for
predicting chemical activity. Information Sciences, Volumes 403–404, September
2017, Pages 42-54. (IF 4.832)
9 / 47
Qualitative Structure-activity
Relationships Approach
• A QSARs are statistical models
correlating one or more piece of
response data about chemicals
(chemical compound biological
activity), with the information
numerically encoded in the form
of descriptors (structural
descriptors).
10 / 47
Qualitative Structure-activity Relationships Approach
Molecular Descriptors
• Molecular descriptors are numerical values that characterize properties
of molecules.
Molecular
descriptors
2.54
• The basic implementation of this topology-based descriptor uses the
information contained in the shortest-distance matrix M.
• Topological descriptors.
• Topological indices are global features that derive information from the
adjacency matrix of a molecular graph.
• An example of Topological descriptors is Wiener Index.
11 / 47
Qualitative Structure-activity Relationships Approach
QSAR For Predicting Biological Activity Of 2,4,6-trisubstituted
1,3,5-triazines
• QSAR model is presented for predicting
biological activity of 2,4,6-trisubstituted
1,3,5-triazines as cannabinoid receptor
(CB2) agonists using random forest (RF),
decision tree (DT), and support vector
machine (SVM).
• The endocannabinoid system plays an
important role in nervous systems.
• A vast variety of diseases and conditions
have been associated with alterations in the
neurotransmitter system.
12 / 47
Qualitative Structure-activity Relationships Approach
Dataset Characteristics And Descriptors
• An investigation to find analogs of
2,4,6-trisubstituted 1,3,5-triazines to
test them as to be CB2 agonist.
• A dataset contains 58 components
and its CB2 agonist activity was
presented.
*Yrj, S., Kalliokoski, T., Laitinen, T., Poso, A., Parkkari, T., Nevalainen, T., Discovery
of novel cannabinoid receptor ligands by a virtual screening approach: Further
development of 2,4,6-trisubstituted 1,3,5-triazines as CB2 agonists. Eur. J. Pharm.
Sci. 48(1-2), 9–20 (2013)
13 / 47
Qualitative Structure-activity Relationships Approach
Model Design
1- Molecular
descriptors
• The twenty molecular descriptors were computed to each chemical
component.
2- Redundant
information
• A pair wise correlation analysis for descriptors
• Only one of any two descriptors with r ≥ 0.97 was excluded to reduce
redundant information.
3- Training
and test
• The remaining descriptors were used to train RF to predicate activity
of each component and to compare result we train DT and SVM for
the same data.
14 / 47
The result shows that random forest has the best prediction accuracy with overall
accuracy equal to 100%
Support vector machine has the worst prediction accuracy with overall accuracy
equal to 67%.
The result also shows that Decision tree has 93% overall accuracy of prediction.
Qualitative Structure-activity Relationships Approach
Results
15 / 47
Qualitative Structure-activity Relationships Approach
Approach Limitation
• Time-consuming
• Computation resources
• Did not give sub-structure
information
• Which descriptors to use?
(Molecular Descriptors for
Chemo-informatics: Volume I
contains more than 3300
descriptors)
16 / 47
Graph Algorithms based-Approach
• New encoding system is presented.
• The encoding system is used to find a relationship
between chemical compounds.
• Presents two new algorithms based on kernel
functions to solve activity prediction problem for
chemical compounds.
• The proposed algorithms were compared with many
other classification methods and the results show
competitive accuracy with these methods.
17 / 47
Graph Algorithms based-Approach
Treelets
• The proposed algorithms were inspired by the enumeration of subtrees of
graphs up to size 6 nodes.
• The resulting patterns are called treelets.
* B. Gaüzère, L. Brun, D. Villemin, Two new graphs kernels in chemoinformatics,
Pattern Recogn. Lett. 33 (15) (2012) 2038–2047.
18 / 47
Graph Algorithms based-Approach
Relation Between Treelets and Stars Shape
One star contains 6
treelets
Path of two stars contains 4
treelets
19 / 47
Graph Algorithms based-Approach
Atoms Coding System
• To encode atoms we create a table
with 5 rows and 118 columns (as the
atoms can be labeled by each of the
118 chemical elements).
• The numbers in the first row represent
central atom type.
• The numbers in the other four rows
represent one type of the chemical
bonds (single, aromatic, double, and
triple bond).
• Numbers in that table are prime
numbers.
20 / 47
Graph Algorithms based-Approach
Coding Example
• The given atom is encoded as the
following:
• Mapping each atom to a
corresponding prime number
from the given table.
• The product of these prime
numbers represents the code of
the atom which equals 2 × 3 ×5
× 17 = 510.
21 / 47
Graph Algorithms based-Approach
Unique Codes
• For any two atoms have the same code, it is
clear that they must be of the same type and
have the same neighbors.
• These atoms not only have the same neighbors
but also they are connected to their neighbors
by the same types of chemical bonds.
22 / 47
Graph Algorithms based-Approach
Paths Of The Feature-coding System
• The path of features is an ordered set of consecutive features.
• Suppose there are n stars in a graph with star codes C1,C2, …,Cn.
• A path of stars of length m in the graph that consists of an ordered set of
star codes, is given by
23 / 47
Graph Algorithms based-Approach
A New Atoms Similarity Algorithm
24 / 47
A New Atoms Similarity Algorithm
Atoms Relationship Matrix
• Matrix S is a squared matrix
which its size equals to the
number of unique star sub-
graphs in dataset.
• Each element at position i, j
in S matrix represents the
number of common
connected sub-graphs
between stars sub-graphs i
and j.
25 / 47
A New Atoms Similarity Algorithm
Union Graph
• Let we have chemical compound gi then it has a corresponding list of
atoms codes Ci, where Ci is defined by Eq. (5):
• For any two chemical compounds gi, gj we define a union atoms codes
list CU
i,j between their atoms codes lists Ci and Cj by Eq. (6):
26 / 47
A New Atoms Similarity Algorithm
Similarity Vector
• We denote by gs
i,j the similarity between two chemical compounds gi and gj
• Similarity vector vs
i for a given chemical compound gi which belongs to a
dataset D of chemical compounds is the vector that contains the similarity
between gi and each gj∈ D.
• Where h is the size of the dataset D.
27 / 47
A New Atoms Similarity Algorithm
Kernel Function
• Two kernel functions are proposed.
• The first one uses the similarity vectors as features for a polynomial kernel
function of order two and we called it the atom similarity kernel Ks.
• The second kernel function uses the outputs of Eq. (7) as kernel matrix entities
and we called it the direct kernel Kd .
28 / 47
Experimental Results
29 / 47
A New Atoms Similarity Algorithm
Datasets Characteristics
• Two experimental evaluations using two
chemical compounds datasets:
• The monoamine oxidase (MAO) dataset
which is composed of 68 molecules : 38
molecules inhibit the monoamine
oxidase (antidepressant drugs) and 30 do
not.
• The AIDS Antiviral Screen Database of
active Compounds and it is composed of
2000 chemical compounds, these
chemical compounds have been screened
as active or inactive against HIV.
MAO AIDS
Size 68 2000
No.
Atoms
(mean)
18.4 15.7
Atoms
Degree
(mean)
2.1 2.1
Smallest 11 2
Largest 27 95
30 / 47
A New Atoms Similarity Algorithm
Prediction Accuracy for (MAO) Dataset
31 / 47
A New Atoms Similarity Algorithm
Prediction Accuracy for (AIDS) Dataset
32 / 47
A New Atoms Similarity Algorithm
Complexity and Limitation
• For one graph the proposed algorithm needs O(nd)
where for treelets kernel it needs O(nd5)
• Accuracy is effected by the number and type of feature.
33 / 47
A New Paths Of Stars
Algorithm
34 / 47
A New Paths Of Stars Algorithm
Path Of Stars Main Algorithm
35 / 47
A New Paths Of Stars Algorithm
Star Relationship Matrices
•The relationship between stars is
represented by two matrices:
• The first matrix is the common sub-stars
matrix denoted by MC.
• The second matrix is the difference sub-
stars matrix denoted by MD.
36 / 47
A New Paths Of Stars Algorithm
Star Relationship Matrices
• The first matrix, MC, represents the greatest common connected
subgraph between two stars.
• The second matrix MD contains the greatest connected subgraph that
exists in Si and not in Sj.
37 / 47
A New Paths Of Stars Algorithm
Star Relationship Matrices Example
MC MD
38 / 47
A New Paths Of Stars Algorithm
Paths of Feature Relationship Matrices
39 / 47
A New Paths Of Stars Algorithm
Paths of Feature Relationship Matrices Example
40 / 47
A New Paths Of Stars Algorithm
Relationship Vectors and Kernel Functions
• Similarity vectors
• Distance vectors
• The second kernel Kdis uses the distance vectors
• The first kernel Ksim uses the similarity vectors
41 / 47
EXPERIMENTAL
RESULTS
42 / 47
A New Paths Of Stars Algorithm
Prediction Accuracy for (MAO) Dataset
43 / 47
A New Paths Of Stars Algorithm
Prediction Accuracy for (AIDS) Dataset
44 / 47
A New Paths Of Stars Algorithm
Features Reduction
MAO
• In the proposed algorithm, only 76
paths of stars are counted.
• The proposed algorithm, which is
combined with distance kernel Kdis, has
an accuracy of 94%.
• This accuracy is similar to the accuracy
of the treelets kernel function, which
count 153 different treelets.
AIDS
• In the proposed algorithm, only 4046
paths of stars are counted.
• The proposed algorithm, which is
combined with distance kernel Kdis,
has an accuracy of 99.5%.
• This accuracy is greater than the
accuracy of the treelets kernel function,
which count 4875 different treelets
45 / 47
A New Paths Of Stars Algorithm
Complexity
• For one graph the proposed algorithm needs O(nd2) where for treelets
kernel it needs O(nd5)
46 / 47
Conclusion & Future Works
• Two approaches were proposed in this thesis:
• The quantitative structure activity relationship (QSAR) approach and it was tested with three
classifiers RF, DT, SVM.
• The approach is tested to predicate biological activity of CB2 receptors agonists and the result shows
that RF has the best prediction accuracy flowed by the DT and the least accuracy was obtained with
SVM.
• The graph algorithms approach
• New Coding System based on primes numbers.
• A new atoms similarity algorithm is proposed with complexity O(nd) and tested on:
• MAO dataset with accuracy 91% and improved by features selection to 98.5%.
• ADIS dataset with accuracy 99.2% and improved by features selection to 99.4%.
• A new paths of stars algorithm is proposed with complexity O(nd2) and tested on:
• MAO dataset with accuracy 94% and reduced the features from 153 to 76.
• ADIS dataset with accuracy 99.5% and reduced the features from 4875 to 4046.
• The proposed algorithms can be further upgraded to improve its performance and
accuracy.
• This modification can be done by using additional types of features in the paths rather than star-
type features only.
• These features can be chosen to be (3D) instead of (2D).
47 / 47
Publications
Two-Class Support Vector Machine with New Kernel
Function Based on Paths of Features for Predicting
Chemical Activity, Information Sciences, Volumes
403–404, September 2017, Pages 42-54 (IF= 4.838 &
SJR = 2.513)
http://www.sciencedirect.com/science/article/pii/S00200
25517306448
Predicting activity approach based on new
atoms similarity kernel function, Journal of
Molecular Graphics and Modelling, Vl. 60,
pp. 55–62, 2015 (IF= 1.674 & SJR=0.480)
http://www.sciencedirect.com/science/ar
ticle/pii/S1093326315300061
48 / 47
ACKNOWLEDGMENT

More Related Content

What's hot

Molecular and Quantum Mechanics in drug design
Molecular and Quantum Mechanics in drug designMolecular and Quantum Mechanics in drug design
Molecular and Quantum Mechanics in drug designAjay Kumar
 
Gaussian presentation
Gaussian presentationGaussian presentation
Gaussian presentationmojdeh y
 
Computational Chemistry: From Theory to Practice
Computational Chemistry: From Theory to PracticeComputational Chemistry: From Theory to Practice
Computational Chemistry: From Theory to PracticeDavid Thompson
 
Software resources in drug design
Software resources in drug designSoftware resources in drug design
Software resources in drug designSurmil Shah
 
STATISTICAL METHOD OF QSAR
STATISTICAL METHOD OF QSARSTATISTICAL METHOD OF QSAR
STATISTICAL METHOD OF QSARRaniBhagat1
 
2.molecular modelling intro
2.molecular modelling intro2.molecular modelling intro
2.molecular modelling introAbhijeet Kadam
 
QSAR statistical methods for drug discovery(pharmacology m.pharm2nd sem)
QSAR statistical methods for drug discovery(pharmacology m.pharm2nd sem)QSAR statistical methods for drug discovery(pharmacology m.pharm2nd sem)
QSAR statistical methods for drug discovery(pharmacology m.pharm2nd sem)Satigayatri
 
review on molecular modeling
review on molecular modelingreview on molecular modeling
review on molecular modelingAnkita Shukla
 
A Genetic Algorithm on Optimization Test Functions
A Genetic Algorithm on Optimization Test FunctionsA Genetic Algorithm on Optimization Test Functions
A Genetic Algorithm on Optimization Test FunctionsIJMERJOURNAL
 
computer aided drug designing and molecular modelling
computer aided drug designing and molecular modellingcomputer aided drug designing and molecular modelling
computer aided drug designing and molecular modellingnehla313
 
Density Functional Theory
Density Functional TheoryDensity Functional Theory
Density Functional Theorykrishslide
 
I041214752
I041214752I041214752
I041214752IOSR-JEN
 

What's hot (18)

Molecular and Quantum Mechanics in drug design
Molecular and Quantum Mechanics in drug designMolecular and Quantum Mechanics in drug design
Molecular and Quantum Mechanics in drug design
 
Gaussian presentation
Gaussian presentationGaussian presentation
Gaussian presentation
 
3D QSAR
3D QSAR3D QSAR
3D QSAR
 
Computational Chemistry: From Theory to Practice
Computational Chemistry: From Theory to PracticeComputational Chemistry: From Theory to Practice
Computational Chemistry: From Theory to Practice
 
Software resources in drug design
Software resources in drug designSoftware resources in drug design
Software resources in drug design
 
Molecular modeling in drug design
Molecular modeling in drug designMolecular modeling in drug design
Molecular modeling in drug design
 
STATISTICAL METHOD OF QSAR
STATISTICAL METHOD OF QSARSTATISTICAL METHOD OF QSAR
STATISTICAL METHOD OF QSAR
 
2.molecular modelling intro
2.molecular modelling intro2.molecular modelling intro
2.molecular modelling intro
 
Characterization Techniques of Metamaterials
Characterization Techniques of Metamaterials Characterization Techniques of Metamaterials
Characterization Techniques of Metamaterials
 
QSAR statistical methods for drug discovery(pharmacology m.pharm2nd sem)
QSAR statistical methods for drug discovery(pharmacology m.pharm2nd sem)QSAR statistical methods for drug discovery(pharmacology m.pharm2nd sem)
QSAR statistical methods for drug discovery(pharmacology m.pharm2nd sem)
 
review on molecular modeling
review on molecular modelingreview on molecular modeling
review on molecular modeling
 
Dv33736740
Dv33736740Dv33736740
Dv33736740
 
Molecular mechanics and dynamics
Molecular mechanics and dynamicsMolecular mechanics and dynamics
Molecular mechanics and dynamics
 
A Genetic Algorithm on Optimization Test Functions
A Genetic Algorithm on Optimization Test FunctionsA Genetic Algorithm on Optimization Test Functions
A Genetic Algorithm on Optimization Test Functions
 
computer aided drug designing and molecular modelling
computer aided drug designing and molecular modellingcomputer aided drug designing and molecular modelling
computer aided drug designing and molecular modelling
 
Density Functional Theory
Density Functional TheoryDensity Functional Theory
Density Functional Theory
 
I041214752
I041214752I041214752
I041214752
 
Molecular modelling
Molecular modelling Molecular modelling
Molecular modelling
 

Similar to Novel algorithms for detection of unknown chemical molecules with specific biological activities

Multiobjective Optimization Tool for a Free Structure Analog Circuits Design ...
Multiobjective Optimization Tool for a Free Structure Analog Circuits Design ...Multiobjective Optimization Tool for a Free Structure Analog Circuits Design ...
Multiobjective Optimization Tool for a Free Structure Analog Circuits Design ...Yaser Kalifa
 
An Automatic Clustering Technique for Optimal Clusters
An Automatic Clustering Technique for Optimal ClustersAn Automatic Clustering Technique for Optimal Clusters
An Automatic Clustering Technique for Optimal ClustersIJCSEA Journal
 
Basics of QSAR Modeling
Basics of QSAR ModelingBasics of QSAR Modeling
Basics of QSAR ModelingPrachi Pradeep
 
Using open bioactivity data for developing machine-learning prediction models...
Using open bioactivity data for developing machine-learning prediction models...Using open bioactivity data for developing machine-learning prediction models...
Using open bioactivity data for developing machine-learning prediction models...Sunghwan Kim
 
A Threshold Fuzzy Entropy Based Feature Selection: Comparative Study
A Threshold Fuzzy Entropy Based Feature Selection:  Comparative StudyA Threshold Fuzzy Entropy Based Feature Selection:  Comparative Study
A Threshold Fuzzy Entropy Based Feature Selection: Comparative StudyIJMER
 
Dimensionality Reduction Evolution and Validation
Dimensionality Reduction Evolution and ValidationDimensionality Reduction Evolution and Validation
Dimensionality Reduction Evolution and Validationiosrjce
 
Retrosynthesis tutorial v2
Retrosynthesis tutorial v2Retrosynthesis tutorial v2
Retrosynthesis tutorial v2Wonjun Jeong
 
Bio inspiring computing and its application in cheminformatics
Bio inspiring computing and its application in cheminformaticsBio inspiring computing and its application in cheminformatics
Bio inspiring computing and its application in cheminformaticsabdelazim Galal
 
Data analysis
Data analysisData analysis
Data analysisamlbinder
 
In-silico structure activity relationship study of toxicity endpoints by QSAR...
In-silico structure activity relationship study of toxicity endpoints by QSAR...In-silico structure activity relationship study of toxicity endpoints by QSAR...
In-silico structure activity relationship study of toxicity endpoints by QSAR...Kamel Mansouri
 
QSAR quantitative structure activity relationship
QSAR quantitative structure activity relationship QSAR quantitative structure activity relationship
QSAR quantitative structure activity relationship ZarlishAttique1
 
Comparison Between Clustering Algorithms for Microarray Data Analysis
Comparison Between Clustering Algorithms for Microarray Data AnalysisComparison Between Clustering Algorithms for Microarray Data Analysis
Comparison Between Clustering Algorithms for Microarray Data AnalysisIOSR Journals
 

Similar to Novel algorithms for detection of unknown chemical molecules with specific biological activities (20)

Multiobjective Optimization Tool for a Free Structure Analog Circuits Design ...
Multiobjective Optimization Tool for a Free Structure Analog Circuits Design ...Multiobjective Optimization Tool for a Free Structure Analog Circuits Design ...
Multiobjective Optimization Tool for a Free Structure Analog Circuits Design ...
 
An Automatic Clustering Technique for Optimal Clusters
An Automatic Clustering Technique for Optimal ClustersAn Automatic Clustering Technique for Optimal Clusters
An Automatic Clustering Technique for Optimal Clusters
 
Basics of QSAR Modeling
Basics of QSAR ModelingBasics of QSAR Modeling
Basics of QSAR Modeling
 
Towards More Reliable 13C and 1H Chemical Shift Prediction: A Systematic Comp...
Towards More Reliable 13C and 1H Chemical Shift Prediction: A Systematic Comp...Towards More Reliable 13C and 1H Chemical Shift Prediction: A Systematic Comp...
Towards More Reliable 13C and 1H Chemical Shift Prediction: A Systematic Comp...
 
Using open bioactivity data for developing machine-learning prediction models...
Using open bioactivity data for developing machine-learning prediction models...Using open bioactivity data for developing machine-learning prediction models...
Using open bioactivity data for developing machine-learning prediction models...
 
A Threshold Fuzzy Entropy Based Feature Selection: Comparative Study
A Threshold Fuzzy Entropy Based Feature Selection:  Comparative StudyA Threshold Fuzzy Entropy Based Feature Selection:  Comparative Study
A Threshold Fuzzy Entropy Based Feature Selection: Comparative Study
 
Dimensionality Reduction Evolution and Validation
Dimensionality Reduction Evolution and ValidationDimensionality Reduction Evolution and Validation
Dimensionality Reduction Evolution and Validation
 
H017365863
H017365863H017365863
H017365863
 
Computational chemistry
Computational chemistryComputational chemistry
Computational chemistry
 
Retrosynthesis tutorial v2
Retrosynthesis tutorial v2Retrosynthesis tutorial v2
Retrosynthesis tutorial v2
 
NMR Chemical Shift Prediction by Atomic Increment-Based Algorithms
NMR Chemical Shift Prediction by Atomic Increment-Based AlgorithmsNMR Chemical Shift Prediction by Atomic Increment-Based Algorithms
NMR Chemical Shift Prediction by Atomic Increment-Based Algorithms
 
Bio inspiring computing and its application in cheminformatics
Bio inspiring computing and its application in cheminformaticsBio inspiring computing and its application in cheminformatics
Bio inspiring computing and its application in cheminformatics
 
Master's Thesis Presentation
Master's Thesis PresentationMaster's Thesis Presentation
Master's Thesis Presentation
 
Data analysis
Data analysisData analysis
Data analysis
 
F04463437
F04463437F04463437
F04463437
 
In-silico structure activity relationship study of toxicity endpoints by QSAR...
In-silico structure activity relationship study of toxicity endpoints by QSAR...In-silico structure activity relationship study of toxicity endpoints by QSAR...
In-silico structure activity relationship study of toxicity endpoints by QSAR...
 
Applications of Computational Quantum Chemistry
Applications of Computational Quantum ChemistryApplications of Computational Quantum Chemistry
Applications of Computational Quantum Chemistry
 
B04302018024
B04302018024B04302018024
B04302018024
 
QSAR quantitative structure activity relationship
QSAR quantitative structure activity relationship QSAR quantitative structure activity relationship
QSAR quantitative structure activity relationship
 
Comparison Between Clustering Algorithms for Microarray Data Analysis
Comparison Between Clustering Algorithms for Microarray Data AnalysisComparison Between Clustering Algorithms for Microarray Data Analysis
Comparison Between Clustering Algorithms for Microarray Data Analysis
 

More from Aboul Ella Hassanien

الأطر والمبادئ الاخلاقية للذكاء الاصطناعي التوليدى.pdf
الأطر والمبادئ الاخلاقية  للذكاء الاصطناعي التوليدى.pdfالأطر والمبادئ الاخلاقية  للذكاء الاصطناعي التوليدى.pdf
الأطر والمبادئ الاخلاقية للذكاء الاصطناعي التوليدى.pdfAboul Ella Hassanien
 
دعوة للاستخدام المسؤول للذكاء الاصطناعي التوليدي في الأوساط الأكاديمية المعر...
دعوة للاستخدام المسؤول للذكاء الاصطناعي التوليدي في الأوساط الأكاديمية  المعر...دعوة للاستخدام المسؤول للذكاء الاصطناعي التوليدي في الأوساط الأكاديمية  المعر...
دعوة للاستخدام المسؤول للذكاء الاصطناعي التوليدي في الأوساط الأكاديمية المعر...Aboul Ella Hassanien
 
حوار مع الأستاذ الدكتور أبو العلا عطيفى حسنين - تقنية الذكاء الاصطناعي تحول م...
حوار مع الأستاذ الدكتور أبو العلا عطيفى حسنين - تقنية الذكاء الاصطناعي تحول م...حوار مع الأستاذ الدكتور أبو العلا عطيفى حسنين - تقنية الذكاء الاصطناعي تحول م...
حوار مع الأستاذ الدكتور أبو العلا عطيفى حسنين - تقنية الذكاء الاصطناعي تحول م...Aboul Ella Hassanien
 
الطاقة من الفضاء: علماء ينقلون الطاقة الشمسية إلى الأرض عن طريق الفضاء لأول م...
الطاقة من الفضاء: علماء ينقلون الطاقة الشمسية إلى الأرض عن طريق الفضاء لأول م...الطاقة من الفضاء: علماء ينقلون الطاقة الشمسية إلى الأرض عن طريق الفضاء لأول م...
الطاقة من الفضاء: علماء ينقلون الطاقة الشمسية إلى الأرض عن طريق الفضاء لأول م...Aboul Ella Hassanien
 
Intelligent Avatars in the Metaverse.pptx
Intelligent Avatars in the Metaverse.pptxIntelligent Avatars in the Metaverse.pptx
Intelligent Avatars in the Metaverse.pptxAboul Ella Hassanien
 
دليل البحث العلمى .pdf
دليل البحث العلمى .pdfدليل البحث العلمى .pdf
دليل البحث العلمى .pdfAboul Ella Hassanien
 
الذكاء الإصطناعى وافاقه فى التعليم على مستوى الوطن العربى: مستوى السياسات
الذكاء الإصطناعى وافاقه فى التعليم على مستوى الوطن العربى: مستوى السياسات الذكاء الإصطناعى وافاقه فى التعليم على مستوى الوطن العربى: مستوى السياسات
الذكاء الإصطناعى وافاقه فى التعليم على مستوى الوطن العربى: مستوى السياسات Aboul Ella Hassanien
 
الصحافة والإعلام الرقمى فى عصر الذكاء الاصطناعي
الصحافة والإعلام الرقمى  فى عصر الذكاء الاصطناعي  الصحافة والإعلام الرقمى  فى عصر الذكاء الاصطناعي
الصحافة والإعلام الرقمى فى عصر الذكاء الاصطناعي Aboul Ella Hassanien
 
الميتافيرس و مستقبل التعليم فى الوطن العربى
الميتافيرس و مستقبل التعليم فى الوطن العربى الميتافيرس و مستقبل التعليم فى الوطن العربى
الميتافيرس و مستقبل التعليم فى الوطن العربى Aboul Ella Hassanien
 
الذكاء الأصطناعى المسؤول ومستقبل الأمن المناخى وانعكاساته الاجتماعية والأمنية
الذكاء الأصطناعى المسؤول ومستقبل  الأمن المناخى وانعكاساته الاجتماعية والأمنيةالذكاء الأصطناعى المسؤول ومستقبل  الأمن المناخى وانعكاساته الاجتماعية والأمنية
الذكاء الأصطناعى المسؤول ومستقبل الأمن المناخى وانعكاساته الاجتماعية والأمنيةAboul Ella Hassanien
 
الذكاء الأصطناعى المسؤول ومستقبل الأمن المناخى وانعكاساته الاجتماعية والأمنية
الذكاء الأصطناعى المسؤول ومستقبل  الأمن المناخى وانعكاساته الاجتماعية والأمنيةالذكاء الأصطناعى المسؤول ومستقبل  الأمن المناخى وانعكاساته الاجتماعية والأمنية
الذكاء الأصطناعى المسؤول ومستقبل الأمن المناخى وانعكاساته الاجتماعية والأمنيةAboul Ella Hassanien
 
التغير المناخى للاطفال
التغير المناخى للاطفالالتغير المناخى للاطفال
التغير المناخى للاطفالAboul Ella Hassanien
 
الذكاء الاصطناعى للاطفال
الذكاء الاصطناعى للاطفالالذكاء الاصطناعى للاطفال
الذكاء الاصطناعى للاطفالAboul Ella Hassanien
 
إستراتيجية مصر للتنمية المستدامة: نحو جائزة الإبتكار والإبداع المؤسسى
إستراتيجية مصر للتنمية المستدامة: نحو جائزة الإبتكار والإبداع المؤسسىإستراتيجية مصر للتنمية المستدامة: نحو جائزة الإبتكار والإبداع المؤسسى
إستراتيجية مصر للتنمية المستدامة: نحو جائزة الإبتكار والإبداع المؤسسىAboul Ella Hassanien
 
الإقتصاد الأخضر لمواجهة التغيرات المناخية
الإقتصاد الأخضر لمواجهة التغيرات المناخية  الإقتصاد الأخضر لمواجهة التغيرات المناخية
الإقتصاد الأخضر لمواجهة التغيرات المناخية Aboul Ella Hassanien
 
الإستخدام المسؤول للذكاء الإصطناعى فى سياق تغيرالمناخ خارطة طريق فى عال...
   الإستخدام المسؤول للذكاء الإصطناعى  فى سياق تغيرالمناخ   خارطة طريق فى عال...   الإستخدام المسؤول للذكاء الإصطناعى  فى سياق تغيرالمناخ   خارطة طريق فى عال...
الإستخدام المسؤول للذكاء الإصطناعى فى سياق تغيرالمناخ خارطة طريق فى عال...Aboul Ella Hassanien
 
الذكاء الإصطناعي والتغيرات المناخية والبيئية:الفرص والتحديات والأدوات السياسية
الذكاء الإصطناعي والتغيرات المناخية والبيئية:الفرص والتحديات والأدوات السياسيةالذكاء الإصطناعي والتغيرات المناخية والبيئية:الفرص والتحديات والأدوات السياسية
الذكاء الإصطناعي والتغيرات المناخية والبيئية:الفرص والتحديات والأدوات السياسيةAboul Ella Hassanien
 
الذكاء الاصطناعى:أسلحة لا تنام وآفاق لا تنتهى
الذكاء الاصطناعى:أسلحة لا تنام وآفاق لا تنتهى الذكاء الاصطناعى:أسلحة لا تنام وآفاق لا تنتهى
الذكاء الاصطناعى:أسلحة لا تنام وآفاق لا تنتهى Aboul Ella Hassanien
 

More from Aboul Ella Hassanien (20)

الأطر والمبادئ الاخلاقية للذكاء الاصطناعي التوليدى.pdf
الأطر والمبادئ الاخلاقية  للذكاء الاصطناعي التوليدى.pdfالأطر والمبادئ الاخلاقية  للذكاء الاصطناعي التوليدى.pdf
الأطر والمبادئ الاخلاقية للذكاء الاصطناعي التوليدى.pdf
 
دعوة للاستخدام المسؤول للذكاء الاصطناعي التوليدي في الأوساط الأكاديمية المعر...
دعوة للاستخدام المسؤول للذكاء الاصطناعي التوليدي في الأوساط الأكاديمية  المعر...دعوة للاستخدام المسؤول للذكاء الاصطناعي التوليدي في الأوساط الأكاديمية  المعر...
دعوة للاستخدام المسؤول للذكاء الاصطناعي التوليدي في الأوساط الأكاديمية المعر...
 
حوار مع الأستاذ الدكتور أبو العلا عطيفى حسنين - تقنية الذكاء الاصطناعي تحول م...
حوار مع الأستاذ الدكتور أبو العلا عطيفى حسنين - تقنية الذكاء الاصطناعي تحول م...حوار مع الأستاذ الدكتور أبو العلا عطيفى حسنين - تقنية الذكاء الاصطناعي تحول م...
حوار مع الأستاذ الدكتور أبو العلا عطيفى حسنين - تقنية الذكاء الاصطناعي تحول م...
 
الطاقة من الفضاء: علماء ينقلون الطاقة الشمسية إلى الأرض عن طريق الفضاء لأول م...
الطاقة من الفضاء: علماء ينقلون الطاقة الشمسية إلى الأرض عن طريق الفضاء لأول م...الطاقة من الفضاء: علماء ينقلون الطاقة الشمسية إلى الأرض عن طريق الفضاء لأول م...
الطاقة من الفضاء: علماء ينقلون الطاقة الشمسية إلى الأرض عن طريق الفضاء لأول م...
 
Intelligent Avatars in the Metaverse.pptx
Intelligent Avatars in the Metaverse.pptxIntelligent Avatars in the Metaverse.pptx
Intelligent Avatars in the Metaverse.pptx
 
دليل البحث العلمى .pdf
دليل البحث العلمى .pdfدليل البحث العلمى .pdf
دليل البحث العلمى .pdf
 
SRGE photo.pdf
SRGE photo.pdfSRGE photo.pdf
SRGE photo.pdf
 
الذكاء الإصطناعى وافاقه فى التعليم على مستوى الوطن العربى: مستوى السياسات
الذكاء الإصطناعى وافاقه فى التعليم على مستوى الوطن العربى: مستوى السياسات الذكاء الإصطناعى وافاقه فى التعليم على مستوى الوطن العربى: مستوى السياسات
الذكاء الإصطناعى وافاقه فى التعليم على مستوى الوطن العربى: مستوى السياسات
 
الصحافة والإعلام الرقمى فى عصر الذكاء الاصطناعي
الصحافة والإعلام الرقمى  فى عصر الذكاء الاصطناعي  الصحافة والإعلام الرقمى  فى عصر الذكاء الاصطناعي
الصحافة والإعلام الرقمى فى عصر الذكاء الاصطناعي
 
الميتافيرس و مستقبل التعليم فى الوطن العربى
الميتافيرس و مستقبل التعليم فى الوطن العربى الميتافيرس و مستقبل التعليم فى الوطن العربى
الميتافيرس و مستقبل التعليم فى الوطن العربى
 
الذكاء الأصطناعى المسؤول ومستقبل الأمن المناخى وانعكاساته الاجتماعية والأمنية
الذكاء الأصطناعى المسؤول ومستقبل  الأمن المناخى وانعكاساته الاجتماعية والأمنيةالذكاء الأصطناعى المسؤول ومستقبل  الأمن المناخى وانعكاساته الاجتماعية والأمنية
الذكاء الأصطناعى المسؤول ومستقبل الأمن المناخى وانعكاساته الاجتماعية والأمنية
 
الذكاء الأصطناعى المسؤول ومستقبل الأمن المناخى وانعكاساته الاجتماعية والأمنية
الذكاء الأصطناعى المسؤول ومستقبل  الأمن المناخى وانعكاساته الاجتماعية والأمنيةالذكاء الأصطناعى المسؤول ومستقبل  الأمن المناخى وانعكاساته الاجتماعية والأمنية
الذكاء الأصطناعى المسؤول ومستقبل الأمن المناخى وانعكاساته الاجتماعية والأمنية
 
التغير المناخى للاطفال
التغير المناخى للاطفالالتغير المناخى للاطفال
التغير المناخى للاطفال
 
الذكاء الاصطناعى للاطفال
الذكاء الاصطناعى للاطفالالذكاء الاصطناعى للاطفال
الذكاء الاصطناعى للاطفال
 
إستراتيجية مصر للتنمية المستدامة: نحو جائزة الإبتكار والإبداع المؤسسى
إستراتيجية مصر للتنمية المستدامة: نحو جائزة الإبتكار والإبداع المؤسسىإستراتيجية مصر للتنمية المستدامة: نحو جائزة الإبتكار والإبداع المؤسسى
إستراتيجية مصر للتنمية المستدامة: نحو جائزة الإبتكار والإبداع المؤسسى
 
الإقتصاد الأخضر لمواجهة التغيرات المناخية
الإقتصاد الأخضر لمواجهة التغيرات المناخية  الإقتصاد الأخضر لمواجهة التغيرات المناخية
الإقتصاد الأخضر لمواجهة التغيرات المناخية
 
الإستخدام المسؤول للذكاء الإصطناعى فى سياق تغيرالمناخ خارطة طريق فى عال...
   الإستخدام المسؤول للذكاء الإصطناعى  فى سياق تغيرالمناخ   خارطة طريق فى عال...   الإستخدام المسؤول للذكاء الإصطناعى  فى سياق تغيرالمناخ   خارطة طريق فى عال...
الإستخدام المسؤول للذكاء الإصطناعى فى سياق تغيرالمناخ خارطة طريق فى عال...
 
الذكاء الإصطناعي والتغيرات المناخية والبيئية:الفرص والتحديات والأدوات السياسية
الذكاء الإصطناعي والتغيرات المناخية والبيئية:الفرص والتحديات والأدوات السياسيةالذكاء الإصطناعي والتغيرات المناخية والبيئية:الفرص والتحديات والأدوات السياسية
الذكاء الإصطناعي والتغيرات المناخية والبيئية:الفرص والتحديات والأدوات السياسية
 
الذكاء الاصطناعى:أسلحة لا تنام وآفاق لا تنتهى
الذكاء الاصطناعى:أسلحة لا تنام وآفاق لا تنتهى الذكاء الاصطناعى:أسلحة لا تنام وآفاق لا تنتهى
الذكاء الاصطناعى:أسلحة لا تنام وآفاق لا تنتهى
 
اقتصاد ميتافيرس
اقتصاد ميتافيرساقتصاد ميتافيرس
اقتصاد ميتافيرس
 

Recently uploaded

APPLICATIONS-AC/DC DRIVES-OPERATING CHARACTERISTICS
APPLICATIONS-AC/DC DRIVES-OPERATING CHARACTERISTICSAPPLICATIONS-AC/DC DRIVES-OPERATING CHARACTERISTICS
APPLICATIONS-AC/DC DRIVES-OPERATING CHARACTERISTICSKurinjimalarL3
 
Model Call Girl in Narela Delhi reach out to us at 🔝8264348440🔝
Model Call Girl in Narela Delhi reach out to us at 🔝8264348440🔝Model Call Girl in Narela Delhi reach out to us at 🔝8264348440🔝
Model Call Girl in Narela Delhi reach out to us at 🔝8264348440🔝soniya singh
 
IVE Industry Focused Event - Defence Sector 2024
IVE Industry Focused Event - Defence Sector 2024IVE Industry Focused Event - Defence Sector 2024
IVE Industry Focused Event - Defence Sector 2024Mark Billinghurst
 
Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...
Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...
Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...Dr.Costas Sachpazis
 
microprocessor 8085 and its interfacing
microprocessor 8085  and its interfacingmicroprocessor 8085  and its interfacing
microprocessor 8085 and its interfacingjaychoudhary37
 
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...Soham Mondal
 
College Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
College Call Girls Nashik Nehal 7001305949 Independent Escort Service NashikCollege Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
College Call Girls Nashik Nehal 7001305949 Independent Escort Service NashikCall Girls in Nagpur High Profile
 
SPICE PARK APR2024 ( 6,793 SPICE Models )
SPICE PARK APR2024 ( 6,793 SPICE Models )SPICE PARK APR2024 ( 6,793 SPICE Models )
SPICE PARK APR2024 ( 6,793 SPICE Models )Tsuyoshi Horigome
 
Call Girls Delhi {Jodhpur} 9711199012 high profile service
Call Girls Delhi {Jodhpur} 9711199012 high profile serviceCall Girls Delhi {Jodhpur} 9711199012 high profile service
Call Girls Delhi {Jodhpur} 9711199012 high profile servicerehmti665
 
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur EscortsHigh Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escortsranjana rawat
 
power system scada applications and uses
power system scada applications and usespower system scada applications and uses
power system scada applications and usesDevarapalliHaritha
 
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130Suhani Kapoor
 
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur EscortsHigh Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur EscortsCall Girls in Nagpur High Profile
 
Sachpazis Costas: Geotechnical Engineering: A student's Perspective Introduction
Sachpazis Costas: Geotechnical Engineering: A student's Perspective IntroductionSachpazis Costas: Geotechnical Engineering: A student's Perspective Introduction
Sachpazis Costas: Geotechnical Engineering: A student's Perspective IntroductionDr.Costas Sachpazis
 
What are the advantages and disadvantages of membrane structures.pptx
What are the advantages and disadvantages of membrane structures.pptxWhat are the advantages and disadvantages of membrane structures.pptx
What are the advantages and disadvantages of membrane structures.pptxwendy cai
 
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort serviceGurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort servicejennyeacort
 
HARMONY IN THE NATURE AND EXISTENCE - Unit-IV
HARMONY IN THE NATURE AND EXISTENCE - Unit-IVHARMONY IN THE NATURE AND EXISTENCE - Unit-IV
HARMONY IN THE NATURE AND EXISTENCE - Unit-IVRajaP95
 
Application of Residue Theorem to evaluate real integrations.pptx
Application of Residue Theorem to evaluate real integrations.pptxApplication of Residue Theorem to evaluate real integrations.pptx
Application of Residue Theorem to evaluate real integrations.pptx959SahilShah
 

Recently uploaded (20)

★ CALL US 9953330565 ( HOT Young Call Girls In Badarpur delhi NCR
★ CALL US 9953330565 ( HOT Young Call Girls In Badarpur delhi NCR★ CALL US 9953330565 ( HOT Young Call Girls In Badarpur delhi NCR
★ CALL US 9953330565 ( HOT Young Call Girls In Badarpur delhi NCR
 
APPLICATIONS-AC/DC DRIVES-OPERATING CHARACTERISTICS
APPLICATIONS-AC/DC DRIVES-OPERATING CHARACTERISTICSAPPLICATIONS-AC/DC DRIVES-OPERATING CHARACTERISTICS
APPLICATIONS-AC/DC DRIVES-OPERATING CHARACTERISTICS
 
Model Call Girl in Narela Delhi reach out to us at 🔝8264348440🔝
Model Call Girl in Narela Delhi reach out to us at 🔝8264348440🔝Model Call Girl in Narela Delhi reach out to us at 🔝8264348440🔝
Model Call Girl in Narela Delhi reach out to us at 🔝8264348440🔝
 
IVE Industry Focused Event - Defence Sector 2024
IVE Industry Focused Event - Defence Sector 2024IVE Industry Focused Event - Defence Sector 2024
IVE Industry Focused Event - Defence Sector 2024
 
Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...
Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...
Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...
 
microprocessor 8085 and its interfacing
microprocessor 8085  and its interfacingmicroprocessor 8085  and its interfacing
microprocessor 8085 and its interfacing
 
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptxExploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
 
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
 
College Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
College Call Girls Nashik Nehal 7001305949 Independent Escort Service NashikCollege Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
College Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
 
SPICE PARK APR2024 ( 6,793 SPICE Models )
SPICE PARK APR2024 ( 6,793 SPICE Models )SPICE PARK APR2024 ( 6,793 SPICE Models )
SPICE PARK APR2024 ( 6,793 SPICE Models )
 
Call Girls Delhi {Jodhpur} 9711199012 high profile service
Call Girls Delhi {Jodhpur} 9711199012 high profile serviceCall Girls Delhi {Jodhpur} 9711199012 high profile service
Call Girls Delhi {Jodhpur} 9711199012 high profile service
 
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur EscortsHigh Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
 
power system scada applications and uses
power system scada applications and usespower system scada applications and uses
power system scada applications and uses
 
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
 
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur EscortsHigh Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
 
Sachpazis Costas: Geotechnical Engineering: A student's Perspective Introduction
Sachpazis Costas: Geotechnical Engineering: A student's Perspective IntroductionSachpazis Costas: Geotechnical Engineering: A student's Perspective Introduction
Sachpazis Costas: Geotechnical Engineering: A student's Perspective Introduction
 
What are the advantages and disadvantages of membrane structures.pptx
What are the advantages and disadvantages of membrane structures.pptxWhat are the advantages and disadvantages of membrane structures.pptx
What are the advantages and disadvantages of membrane structures.pptx
 
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort serviceGurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
 
HARMONY IN THE NATURE AND EXISTENCE - Unit-IV
HARMONY IN THE NATURE AND EXISTENCE - Unit-IVHARMONY IN THE NATURE AND EXISTENCE - Unit-IV
HARMONY IN THE NATURE AND EXISTENCE - Unit-IV
 
Application of Residue Theorem to evaluate real integrations.pptx
Application of Residue Theorem to evaluate real integrations.pptxApplication of Residue Theorem to evaluate real integrations.pptx
Application of Residue Theorem to evaluate real integrations.pptx
 

Novel algorithms for detection of unknown chemical molecules with specific biological activities

  • 1. 1 / 47 NOVELALGORITHMS FOR DETECTION OF UNKNOWN CHEMICAL MOLECULES WITH SPECIFIC BIOLOGICAL ACTIVITIES By Ahmed Hassan Ahmed Supervisors Prof. Dr. Effat A. Saied Prof. Dr. Aboul Ella Hassanien
  • 2. 2 / 47 Agenda • Introduction – Chemo-informatics – Drug discovery and its cyclic – Molecular representation • Problem Statement • The proposed approaches • Qualitative structure-activity relationships approach • Graph algorithms based-approach • Coding System • A new atoms similarity algorithm • A new paths of stars algorithm • Conclusion and future work
  • 3. 3 / 47 Introduction Chemo-informatics Chemistry Problems Information Problems Cheminformatics is the use of computer and informational techniques applied to a range of problems in the field of chemistry. Chemoinformatics / cheminformatics can be broadly defined as the field of solving chemical problems with computers.
  • 4. 4 / 47 Introduction Drug Discovery and its cycle • Drug discovery is the process through which potential new medicines are identified. https://en.wikipedia.org/wiki/Drug_discovery
  • 5. 5 / 47 Introduction Molecular Activity and Similarity • Similarity principle: • Structurally similar molecules tend to have similar properties • Properties: biological activity, solubility, color and so on.
  • 6. 6 / 47 Introduction Molecular Representation • The structure of a molecule can be represented by a graph • Graph = collection of nodes and edges, nodes and edges have properties (atomic number, bond order). • The molecular graph can be presented by Connection table which describes the relationships between nodes.
  • 7. 7 / 47 Problem Statement • The development of a new chemical drug is still a challenging, time-consuming and costly process. • This thesis aims to provide new approaches to reduce the time and the cost of chemical compounds classification. Thousands of Candidate Compounds Hundreds of Candidate Compounds Approved Drug
  • 8. 8 / 47 The Proposed Approaches • Molecular Descriptors and Quantitative-structure activity relationships (QSAR-Model) • Predicting biological activity of 2,4,6-trisubstituted 1,3,5-triazines using random forest, in: Proceedings of the 5th International Conference on Innovations in Bio- Inspired Computing and Applications (IBICA 2014), 2014, pp. 101–110. • Graph Algorithms • A new atoms similarity algorithm • Predicting activity approach based on new atoms similarity kernel function. Journal of Molecular Graphics and Modelling, Volume 60, July 2015, Pages 55 – 62. (IF 1.754) • A new paths of stars algorithm • Two-class support vector machine with new kernel function based on paths of features for predicting chemical activity. Information Sciences, Volumes 403–404, September 2017, Pages 42-54. (IF 4.832)
  • 9. 9 / 47 Qualitative Structure-activity Relationships Approach • A QSARs are statistical models correlating one or more piece of response data about chemicals (chemical compound biological activity), with the information numerically encoded in the form of descriptors (structural descriptors).
  • 10. 10 / 47 Qualitative Structure-activity Relationships Approach Molecular Descriptors • Molecular descriptors are numerical values that characterize properties of molecules. Molecular descriptors 2.54 • The basic implementation of this topology-based descriptor uses the information contained in the shortest-distance matrix M. • Topological descriptors. • Topological indices are global features that derive information from the adjacency matrix of a molecular graph. • An example of Topological descriptors is Wiener Index.
  • 11. 11 / 47 Qualitative Structure-activity Relationships Approach QSAR For Predicting Biological Activity Of 2,4,6-trisubstituted 1,3,5-triazines • QSAR model is presented for predicting biological activity of 2,4,6-trisubstituted 1,3,5-triazines as cannabinoid receptor (CB2) agonists using random forest (RF), decision tree (DT), and support vector machine (SVM). • The endocannabinoid system plays an important role in nervous systems. • A vast variety of diseases and conditions have been associated with alterations in the neurotransmitter system.
  • 12. 12 / 47 Qualitative Structure-activity Relationships Approach Dataset Characteristics And Descriptors • An investigation to find analogs of 2,4,6-trisubstituted 1,3,5-triazines to test them as to be CB2 agonist. • A dataset contains 58 components and its CB2 agonist activity was presented. *Yrj, S., Kalliokoski, T., Laitinen, T., Poso, A., Parkkari, T., Nevalainen, T., Discovery of novel cannabinoid receptor ligands by a virtual screening approach: Further development of 2,4,6-trisubstituted 1,3,5-triazines as CB2 agonists. Eur. J. Pharm. Sci. 48(1-2), 9–20 (2013)
  • 13. 13 / 47 Qualitative Structure-activity Relationships Approach Model Design 1- Molecular descriptors • The twenty molecular descriptors were computed to each chemical component. 2- Redundant information • A pair wise correlation analysis for descriptors • Only one of any two descriptors with r ≥ 0.97 was excluded to reduce redundant information. 3- Training and test • The remaining descriptors were used to train RF to predicate activity of each component and to compare result we train DT and SVM for the same data.
  • 14. 14 / 47 The result shows that random forest has the best prediction accuracy with overall accuracy equal to 100% Support vector machine has the worst prediction accuracy with overall accuracy equal to 67%. The result also shows that Decision tree has 93% overall accuracy of prediction. Qualitative Structure-activity Relationships Approach Results
  • 15. 15 / 47 Qualitative Structure-activity Relationships Approach Approach Limitation • Time-consuming • Computation resources • Did not give sub-structure information • Which descriptors to use? (Molecular Descriptors for Chemo-informatics: Volume I contains more than 3300 descriptors)
  • 16. 16 / 47 Graph Algorithms based-Approach • New encoding system is presented. • The encoding system is used to find a relationship between chemical compounds. • Presents two new algorithms based on kernel functions to solve activity prediction problem for chemical compounds. • The proposed algorithms were compared with many other classification methods and the results show competitive accuracy with these methods.
  • 17. 17 / 47 Graph Algorithms based-Approach Treelets • The proposed algorithms were inspired by the enumeration of subtrees of graphs up to size 6 nodes. • The resulting patterns are called treelets. * B. Gaüzère, L. Brun, D. Villemin, Two new graphs kernels in chemoinformatics, Pattern Recogn. Lett. 33 (15) (2012) 2038–2047.
  • 18. 18 / 47 Graph Algorithms based-Approach Relation Between Treelets and Stars Shape One star contains 6 treelets Path of two stars contains 4 treelets
  • 19. 19 / 47 Graph Algorithms based-Approach Atoms Coding System • To encode atoms we create a table with 5 rows and 118 columns (as the atoms can be labeled by each of the 118 chemical elements). • The numbers in the first row represent central atom type. • The numbers in the other four rows represent one type of the chemical bonds (single, aromatic, double, and triple bond). • Numbers in that table are prime numbers.
  • 20. 20 / 47 Graph Algorithms based-Approach Coding Example • The given atom is encoded as the following: • Mapping each atom to a corresponding prime number from the given table. • The product of these prime numbers represents the code of the atom which equals 2 × 3 ×5 × 17 = 510.
  • 21. 21 / 47 Graph Algorithms based-Approach Unique Codes • For any two atoms have the same code, it is clear that they must be of the same type and have the same neighbors. • These atoms not only have the same neighbors but also they are connected to their neighbors by the same types of chemical bonds.
  • 22. 22 / 47 Graph Algorithms based-Approach Paths Of The Feature-coding System • The path of features is an ordered set of consecutive features. • Suppose there are n stars in a graph with star codes C1,C2, …,Cn. • A path of stars of length m in the graph that consists of an ordered set of star codes, is given by
  • 23. 23 / 47 Graph Algorithms based-Approach A New Atoms Similarity Algorithm
  • 24. 24 / 47 A New Atoms Similarity Algorithm Atoms Relationship Matrix • Matrix S is a squared matrix which its size equals to the number of unique star sub- graphs in dataset. • Each element at position i, j in S matrix represents the number of common connected sub-graphs between stars sub-graphs i and j.
  • 25. 25 / 47 A New Atoms Similarity Algorithm Union Graph • Let we have chemical compound gi then it has a corresponding list of atoms codes Ci, where Ci is defined by Eq. (5): • For any two chemical compounds gi, gj we define a union atoms codes list CU i,j between their atoms codes lists Ci and Cj by Eq. (6):
  • 26. 26 / 47 A New Atoms Similarity Algorithm Similarity Vector • We denote by gs i,j the similarity between two chemical compounds gi and gj • Similarity vector vs i for a given chemical compound gi which belongs to a dataset D of chemical compounds is the vector that contains the similarity between gi and each gj∈ D. • Where h is the size of the dataset D.
  • 27. 27 / 47 A New Atoms Similarity Algorithm Kernel Function • Two kernel functions are proposed. • The first one uses the similarity vectors as features for a polynomial kernel function of order two and we called it the atom similarity kernel Ks. • The second kernel function uses the outputs of Eq. (7) as kernel matrix entities and we called it the direct kernel Kd .
  • 29. 29 / 47 A New Atoms Similarity Algorithm Datasets Characteristics • Two experimental evaluations using two chemical compounds datasets: • The monoamine oxidase (MAO) dataset which is composed of 68 molecules : 38 molecules inhibit the monoamine oxidase (antidepressant drugs) and 30 do not. • The AIDS Antiviral Screen Database of active Compounds and it is composed of 2000 chemical compounds, these chemical compounds have been screened as active or inactive against HIV. MAO AIDS Size 68 2000 No. Atoms (mean) 18.4 15.7 Atoms Degree (mean) 2.1 2.1 Smallest 11 2 Largest 27 95
  • 30. 30 / 47 A New Atoms Similarity Algorithm Prediction Accuracy for (MAO) Dataset
  • 31. 31 / 47 A New Atoms Similarity Algorithm Prediction Accuracy for (AIDS) Dataset
  • 32. 32 / 47 A New Atoms Similarity Algorithm Complexity and Limitation • For one graph the proposed algorithm needs O(nd) where for treelets kernel it needs O(nd5) • Accuracy is effected by the number and type of feature.
  • 33. 33 / 47 A New Paths Of Stars Algorithm
  • 34. 34 / 47 A New Paths Of Stars Algorithm Path Of Stars Main Algorithm
  • 35. 35 / 47 A New Paths Of Stars Algorithm Star Relationship Matrices •The relationship between stars is represented by two matrices: • The first matrix is the common sub-stars matrix denoted by MC. • The second matrix is the difference sub- stars matrix denoted by MD.
  • 36. 36 / 47 A New Paths Of Stars Algorithm Star Relationship Matrices • The first matrix, MC, represents the greatest common connected subgraph between two stars. • The second matrix MD contains the greatest connected subgraph that exists in Si and not in Sj.
  • 37. 37 / 47 A New Paths Of Stars Algorithm Star Relationship Matrices Example MC MD
  • 38. 38 / 47 A New Paths Of Stars Algorithm Paths of Feature Relationship Matrices
  • 39. 39 / 47 A New Paths Of Stars Algorithm Paths of Feature Relationship Matrices Example
  • 40. 40 / 47 A New Paths Of Stars Algorithm Relationship Vectors and Kernel Functions • Similarity vectors • Distance vectors • The second kernel Kdis uses the distance vectors • The first kernel Ksim uses the similarity vectors
  • 42. 42 / 47 A New Paths Of Stars Algorithm Prediction Accuracy for (MAO) Dataset
  • 43. 43 / 47 A New Paths Of Stars Algorithm Prediction Accuracy for (AIDS) Dataset
  • 44. 44 / 47 A New Paths Of Stars Algorithm Features Reduction MAO • In the proposed algorithm, only 76 paths of stars are counted. • The proposed algorithm, which is combined with distance kernel Kdis, has an accuracy of 94%. • This accuracy is similar to the accuracy of the treelets kernel function, which count 153 different treelets. AIDS • In the proposed algorithm, only 4046 paths of stars are counted. • The proposed algorithm, which is combined with distance kernel Kdis, has an accuracy of 99.5%. • This accuracy is greater than the accuracy of the treelets kernel function, which count 4875 different treelets
  • 45. 45 / 47 A New Paths Of Stars Algorithm Complexity • For one graph the proposed algorithm needs O(nd2) where for treelets kernel it needs O(nd5)
  • 46. 46 / 47 Conclusion & Future Works • Two approaches were proposed in this thesis: • The quantitative structure activity relationship (QSAR) approach and it was tested with three classifiers RF, DT, SVM. • The approach is tested to predicate biological activity of CB2 receptors agonists and the result shows that RF has the best prediction accuracy flowed by the DT and the least accuracy was obtained with SVM. • The graph algorithms approach • New Coding System based on primes numbers. • A new atoms similarity algorithm is proposed with complexity O(nd) and tested on: • MAO dataset with accuracy 91% and improved by features selection to 98.5%. • ADIS dataset with accuracy 99.2% and improved by features selection to 99.4%. • A new paths of stars algorithm is proposed with complexity O(nd2) and tested on: • MAO dataset with accuracy 94% and reduced the features from 153 to 76. • ADIS dataset with accuracy 99.5% and reduced the features from 4875 to 4046. • The proposed algorithms can be further upgraded to improve its performance and accuracy. • This modification can be done by using additional types of features in the paths rather than star- type features only. • These features can be chosen to be (3D) instead of (2D).
  • 47. 47 / 47 Publications Two-Class Support Vector Machine with New Kernel Function Based on Paths of Features for Predicting Chemical Activity, Information Sciences, Volumes 403–404, September 2017, Pages 42-54 (IF= 4.838 & SJR = 2.513) http://www.sciencedirect.com/science/article/pii/S00200 25517306448 Predicting activity approach based on new atoms similarity kernel function, Journal of Molecular Graphics and Modelling, Vl. 60, pp. 55–62, 2015 (IF= 1.674 & SJR=0.480) http://www.sciencedirect.com/science/ar ticle/pii/S1093326315300061