Large volume of Genomics data is produced on daily basis due to the advancement in sequencing technology. This data is of no value if it is not properly analysed. Different kinds of analytics are required to extract useful information from this raw data. Classification, Prediction, Clustering and Pattern Extraction are useful techniques of data mining. These techniques require appropriate selection of attributes of data for getting accurate results. However, Bioinformatics data is high dimensional, usually having hundreds of attributes. Such large a number of attributes affect the performance of machine learning algorithms used for classification/prediction. So, dimensionality reduction techniques are required to reduce the number of attributes that can be further used for analysis. In this paper, Principal Component Analysis and Factor Analysis are used for dimensionality reduction of Bioinformatics data. These techniques were applied on Leukaemia data set and the number of attributes was reduced from to.
Deep Dive into Using PCA and Factor Analysis for Dimensionality Reduction of Bio-informatics Data.
Large volume of Genomics data is produced on daily basis due to the advancement in sequencing technology. This data is of no value if it is not properly analysed. Different kinds of analytics are required to extract useful information from this raw data. Classification, Prediction, Clustering and Pattern Extraction are useful techniques of data mining. These techniques require appropriate selection of attributes of data for getting accurate results. However, Bioinformatics data is high dimensional, usually having hundreds of attributes. Such large a number of attributes affect the performance of machine learning algorithms used for classification/prediction. So, dimensionality reduction techniques are required to reduce the number of attributes that can be further used for analysis. In this paper, Principal Component Analysis and Factor Analysis are used for dimensionality reduction of Bioinformatics data. These techniques were applied on Leukaemia data set and the number of attributes
(IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 8, No. 5, 2017
415 | P a g e
www.ijacsa.thesai.org
Using PCA and Factor Analysis for Dimensionality
Reduction of Bio-informatics Data
M. Usman Ali
Department of Computer Science
COMSATS Institute of Information
Technology
Sahiwal, Pakistan
Shahzad Ahmed
Department of Computer Science
COMSATS Institute of Information
Technology
Sahiwal, Pakistan
Javed Ferzund
Department of Computer Science
COMSATS Institute of Information
Technology
Sahiwal, Pakistan
Atif Mehmood
Riphah Institute of Computing and Applied Sciences (RICAS)
Riphah International University
Lahore, Pakistan
Abbas Rehman
Department of Computer Science
COMSATS Institute of Information Technology
Sahiwal, Pakistan
Abstract—Large volume of Genomics data is produced on
daily basis due to the advancement in sequencing technology.
This data is of no value if it is not properly analysed. Different
kinds of analytics are required to extract useful information from
this raw data. Classification, Prediction, Clustering and Pattern
Extraction are useful techniques of data mining. These
techniques require appropriate selection of attributes of data for
getting accurate results. However, Bioinformatics data is high
dimensional, usually having hundreds of attributes. Such large a
number of attributes affect the performance of machine learning
algorithms used for classification/prediction. So, dimensionality
reduction techniques are required to reduce the number of
attributes that can be further used for analysis. In this paper,
Principal Component Analysis and Factor Analysis are used for
dimensionality
reduction
of
Bioinformatics
data.
These
techniques were applied on Leukaemia data set and the number
of attributes was reduced from to.
Keywords—Bioinformatics; Statistics; Microarray; Leukaemia;
Feature Selection; Statistical tests; PCA; Factor Analysis; R tool
I.
INTRODUCTION
Bioinformatics experiments are based on Genome, DNA,
RNA and Chromosomes. Genomics plays an imperative role in
this field. Huge amount of data has been produced in Genomics
with a substantial portion produced in Functional Genomics (in
the form of protein-protein association), Structural Genomics
(in the form of 3-D structure). By using NGS (Next Generation
Sequencing) technique, a lot of work has been done in the field
of Microarray. This technique is helpful to identify human
diseases. NGS is sequencing technique which is used to detect
sequences of proteomics for next generation.
Genetic Diseases are caused by Genetic disorders that are
more complex because of multiple genes interaction. These
disorders are breast cancer, colon cancer, skin cancer, autism,
progeria, and haemophilia. These are caused by mutation in
genes or sometimes inherit from parents. Leukaemia is a
cancer of blood cells that occurs due to genome abnormality.
Microarray includes genes expression data that is present at
large scale.
Bioinformatics data needs to be store in an efficient manner
and include a lot of Attributes (Variables). The major problem
is that, most of tools crash when large data stored in it.
Statistics
plays
superlative
role
in
the
field
of
Bioinformatics, Mathematics and Computer Science. It is used
to extract, organise, analyse and visualise large amount of data.
. For this purpose, a lot of tools like Excel, Weka, Matlab and
R are available. Many Statistical tests are used for the
extraction of relevant information. These are t-test, chi-squared
test (χ2-test), ANOVA (Analysis of Variance), Kruskal-Wallis,
Friedman and PCA (Principle Component Analysis) tests [1]
Statistical t-test is used to check the difference between sample
Mean and hypothesised value. ANOVA is parametric
(distribution) test used to check the difference of dependent
variables with levels of independent variables. Kruskal-Wallis
is non-parametric (distribution free) test in which assumptions
are not including unlike ANOVA. Friedman test is used when
there is one distributed dependent variable and one
independent variable with two or many levels. It is used to
check the difference in reading and math scores and writing.
Chi-squared test is used to compare the observed data with data
according to specific hypothesis. PCA test is used to
Select/Extract relevant and specific information about variables
(attributes) in large dataset. In PCA, correlation is found
between principal components and original data. All of these
tests applied in Statistical tool.
R is open source Statistical tool that is used for loading,
extracting, interpretation and analysis of data. It includes many
operations such as standard deviation, correlation, mean,
variance, median, mode, graphs, plot, charts, and histograms. It
automatically loads libraries and packages. It performs
Machine Learning Classification and Clustering tasks very
quickly and effective
…(Full text truncated)…
This content is AI-processed based on ArXiv data.
