Efficiency of Using Sequence Discovery for Polymorphism in DNA Sequence

95
S. Kalaiselvi and A. Meena, ―Efficiency of using sequence discovery for polymorphism in DNA sequence,‖ International Journal of
Scientific and Technical Advancements, Volume 2, Issue 1, pp. 95-100, 2016.
International Journal of Scientific and Technical Advancements
ISSN: 2454-1532
Efficiency of Using Sequence Discovery for
Polymorphism in DNA Sequence
S. Kalaiselvi1
, A. Meena2
1, 2
Computer Science, Dr. SNS. Rajalakshmi CAS, Coimbatore, Tamil Nadu, India-641049
Email address: 1
kulandaikalai@gmail.com,2
mithra1710@gmail.com
Abstract- DNA-deoxyribonucleic acid, sequencing provide method for efficient large-scale discovery of markers for use in
Human Genes. Discovery options include large-scale gene-enriched genome sequencing and whole genome sequencing.
Sequencing of the transcriptome of genotypes varying for identify genes with patterns of expression that could explain the
phenotypic variation. This thesis presents a effective sequence pattern discovery to find polymorphic motifs in human DNA
sequence to detect hidden genetic medical conditions using three structure LDK Algorithms, HVS and MDFS. To find out the
disease of Ebola sequence pattern matching with polymorphism DNA sequence. The position scrolling matrix used for find
report of test result of sequence pattern.
Keywords—LDK-length distance K-support; HVS-hierarchical verification; MDFS-matching depth first search.
I. INTRODUCTION
A. Overview of Data Mining
Data mining means ―mining‖ knowledge from large data
set. It is defined as ―the process of discovering meaningful
new inter relationship, patterns and mode by finding into large
amounts of data stored in a data set‖.
a) Sequence Mining: Sequence Mining means finding
sequential patterns among the large dataset. It finds out
frequent substring as patterns from a dataset. With
massive amounts of data continuously being gathered,
many industries are becoming interested in digging
sequential patterns from their databases.
b) The goal of sequential data mining:The approach is to
discover frequently occurring patterns but not identical.
The challenge in discovering such patterns is to allow for
some noise in the matching process. To find such a
method first is to find the definition of a pattern, and then
definition of similarity between two patterns. This
similarity definition of the two patterns can vary from one
application to another.
c) DNA sequence mining: DNA sequence is an important
mean to study the structure and function of the DNA
sequence. In this thesis, based on the characteristics of the
DNA sequence an algorithm will proposed which uses the
maximal frequent pattern segments based on adjacent
maximal frequent. DNA sequences use an alphabet {A, C,
G, T} representing the four nitrogenous bases Adenine,
Cytosine, Guanine and Thymine.
DNA sequencing, technique used to determine the
nucleotide sequence of DNA (deoxyribonucleic acid).
B. Human DNA
Every human has his / her unique genes. Genes are made
up of DNA and therefore the DNA sequence of each human is
unique. However the DNA sequences of all humans are 99.9%
identical, which means there is only 0.1% difference. It
contains genetic instructions of an organism and is mainly
composed of nucleotides of four types. Adenine (A), Cytosine
(C), Guanine (G), and Thymine (T). The pattern itself may not
be exactly known, because it may involve insertion, deletion,
or replacement of the symbols.
Fig. 1. Structure of the DNA.
When we know a particular sequence is the cause for a
disease, the trace of the sequence in the DNA and the number
of occurrences of the sequence defines the intensity of the
disease. As the DNA is a large database we need an efficient
algorithm to find out a particular sequence in the given DNA.
We have to find the number of repetitions and the start index
and end index of the sequence, which can be used for the
diagnosis of the disease and also the intensity of the disease by
counting the number of pattern matching strings, occurred in a
gene database.
C. Biological Motivation for Pattern Discovery
Nucleotide and protein sequences contain patterns or
motifs that have been preserved through evolution because

96
ISSN: 2454-1532
they are important to the structure or function of the molecule.
In proteins, these conserved sequences may be involved in the
binding of the protein to its substrate or to another protein,
may comprise the active site of an enzyme or may determine
the three dimensional structure of the protein. Discovery of
motifs in protein and nucleotide sequences can lead to
Determine of function and to elucidation of evolution
relationship among sequences.
D. Objective of the Research
The main aim of this system is to Discovering frequent
approximate sequences selecting randomly one subsequence
of length W from each input sequences. This thesis proposed a
method for model (L, d, k) and tried to improve the
performance of existing algorithm. Some of the existing
methods Hierarchical Verification Algorithm (HVA) and
Matching with Depth First Search (MDFS).The number of
comparisons per character (CPC ratio) which is equal to
(Number of comparisons/file size) can be used as a
measurement factor. The main objective of this this work is to
identify the motif sequence pattern of Ebola virus infected
patients DNA Sequence to determine the immunity resistant.
E. Scope of the Research
The Scope of the system is identify a genotyping platform
and molecular marker that are amenable to high-throughput
genotyping for EBOLA Virus. This approach explores
alternative approaches for virus infection identification, using
single nucleotide polymorphism (SNP) molecular markers and
high throughput genotyping platforms which reduce the cost
per data point. SNPs are unambiguous, genome-wide
molecular markers that allow identification and traceability
through-out the production chain.
F. Problem Definition
Sequential pattern mining deals with data in large
sequential data sets. The results of pattern mining can be used
for planning and prediction. The difficulty in discovering
frequent patterns is to allow for some noise in the matching
process. The most important part of pattern discovery is the
definition of a pattern and similarity between two patterns
which may vary from one application to another. Discovering
frequent patterns is computationally expensive process and
counting the instances of pattern requires a large amount of
processing time. These algorithms differ in their ways of
traversing the item set lattice, dimension and the way in which
they handle the database; i.e., how many passes they make
over the entire database and how they reduce the size of the
processed database in each passes.
II. LITERATURE REVIEW
A. Medical Research
Most current successful applications of data mining in
Health Informatics are in the subfield of medical research. The
reason is that most of the current health related data are stored
in small datasets scattered through various clinics, hospitals,
and research centers. However, most applications of data
mining in clinical and administrative decision support systems
require homogeneous and centralized data. On the other hand,
data mining methods can still be successfully applied on small
and scattered datasets, and help researchers extract insightful
patterns, cause and effect relationships, and predictive scoring
systems from currently available data.
B. Association Rules and Decision Trees
Association rules and decision trees for disease prediction
Ordonez applies different classifiers, associative classifier and
decision trees, for predicting the percentage of vessel
narrowing (LDA, RCA, LCX and LM) compare to a healthy
artery [15]. The dataset contains 655 patient records with 25
medical attributes. Three main issues about mining associative
rules in medical datasets are mentioned in this work. A
significant fraction of association rules are irrelevant and most
relevant rules with high quality metrics appear only at low
support. On the other hand, the number of discovered rules
becomes extremely large at low support. Hence, association
rules are used with constraints. Each item corresponds to the
presence or absence of one categorical value or one numeric
interval. First constraint is that there is a limit on the
maximum item-set size. Second, the items are grouped and in
each association, there is at most one from each group. The
third constraint is that each item can only appear in antecedent
or consequent. The result from associative classifier is
compared with two decision tree algorithms CN4.5 and
CART. The authors demonstrate that associative rules can do
better than decision trees for predicting diseased arteries.
C. Apriori-Based Algorithms
The Apriori and Apriori All set the basis for a breed of
algorithms that depend largely on the Apriori property and use
the Apriori-generate join procedure to generate candidate
sequences. The Apriori property states that All non-empty
subsets of a frequent item set must also be frequent. It is also
described as anti-monotonic, in that if a sequence cannot pass
of these candidates. Frequent sequence produced from these
candidates are captured, while those candidates without
minimum support are removed .This procedure is repeated
until all the candidates have been counted. In the first step
GSP algorithm [2] finds all the length-1 candidates (using one
database scan) and orders them with respect to their support
ignoring ones for which support (<min_sup). Then for each
level, this algorithm scans datasets to collect support count for
each candidate sequence and generates candidate length (k+1)
sequences from length-k frequent sequences using Apriori
algorithm. This is repeated until frequent sequence or no
candidate can be found.
D. GSP
The GSP algorithm describe by Agrawal and Shrikant [1]
makes multiple passes over the data. The algorithm not a main
memory algorithms. If the candidates do not fit memory, the
algorithm generate only the many candidates as will fit in
memory and the date is scanned to count the support of these
candidates. Frequent sequence produced from these candidates
are captured, while those candidates without Frequent
sequence produced from these candidates are captured, while

97
ISSN: 2454-1532
those candidates without minimum support are removed. This
procedure is repeated until all the candidates have been
counted. In the first step GSP algorithm [2] finds all the
length-1 candidates (using one database scan) and orders them
with respect to their support the minimum support threshold,
its entire super sequences will never pass the test.
E. Spade
Besides the horizontal formatting method (GSP) [2], the
sequence database can be transformed into a vertical format
consisting of items’ id-lists., is a list of (sequence-id, time
stamp) pairs indicating the occurring time stamps of the item
in that sequence. Searching in the lattice of the datasets
formed by id-list intersections, the SPADE [3] (Sequential
Pattern Discovery using Equivalence classes) algorithm
presented by M.J.Jaki completes the mining in three passes of
scanning the databases. However, additional computation time
is required to transform this database of horizontal layout to
vertical layout, which also requires additional storage space
which is larger than that of the original sequence database.
F. Pattern Growth Based Algorithms
In pattern growth algorithm partitioning of search space
feature plays a major role. In each pattern growth algorithm
working starts by representing the database sequence and then
provides the way to partition the database search space and
thus producing as less candidate patterns as possible by
growing on the already mined frequent sequences, and
applying the Apriori algorithm as the search space is being
traversed recursively looking for frequent sequences.
(a) Search space partitioning: It allows the generated search
space having candidate sequence to be partitioned to
make the efficient use of memory. Several ways are
available for the partitioning of the search space. Among
which first is to partition the search space in smaller
blocks in parallel. Modern techniques include projected
and conditional search.
(b) Tree projection: In this algorithms implement a physical
tree data structure representation of the search space,
which searches the frequent patterns either using breadth
first or depth first search and an Apriori algorithm is used
for minimizing the sequence length.
(c) Depth-first traversal: That depth-first search of the search
space makes a big difference in performance, and also
helps in the early pruning of candidate sequences as well
as mining of closed sequences. The fact that depth-first
traversal used is that utilizes less memory, more directed
search space, and thus less candidate sequence generation
than breadth-first algorithms.
G. Application of Data Mining Techniques In CRM
Data mining technique is used in CRM .Now a days it is
one of the hot topic to research in the industry because CRM
have attracted both the practitioners and academics. It aims to
give a research summary on the application of data mining in
the CRM domain and techniques, which are most often used.
Research on the application of data mining in CRM will
increase significantly in the future based on past publication
rates and the increasing interest in the area.
III. RESEARCH METHODOLOGY
A. DNA Input Sequences Representation
The input of pattern discovery programs usually consists of
several sequences, expected to contain the pattern. We will
denote Σ the alphabet of all possible characters occurring in
the sequences. Thus Σ = {A,C,G,T} for DNA sequences and Σ
is a set of all 20 amino acids for protein sequences.
 Ambiguous character is a character corresponding to a
subset of Σ. Ambiguous character then matches any
character from this set. Such sets are usually denoted by a
list of its members enclosed in square brackets e.g. [LF] is
a set containing L and F. A-[LF]-G is a pattern in a
notation used in EBOLA database. This patterns matches
3-character subsequences starting with A, ending with G
and having either L or F in the middle. For nucleotide
sequence there is a special letter for each set of
nucleotides, where R=[AG], Y=[CT], W=[AT], S=[GC],
B=[CGT], D=[AGT], H=[ACT], V=[ACG], N=[ACGT].
 Wild-card or don’t care is a special kind of ambiguous
character that matches any character from Σ. Wild-cards
are denoted N in nucleotide sequences, X in protein
sequences. Often they are also denoted by dot ’.’.
Sequence of one or several consecutive wild-cards is called
gap and patterns allowing wild-cards are often called
gapped patterns.
 Flexible gap is a gap of variable length. In EBOLA
database it is denoted by x(i,j) where i is the lower bound
on the gap length and j is an upper bound. Thus x(4,6)
matches any gap with length 4, 5, or 6. They also denote a
fixed gap of length i as x(i) (e.g. x(3) = ...). Finally (*)
denotes gap of any length (possibly 0).
 Following string is an example of a EBOLA pattern
containing all mentioned features: [F-x(5)-G-x(2,4)-G-*-
H]. Some programs do not allow all these features, for
example they do not allow flexible gaps or they allow any
gaps but do not allow ambiguous characters other than a
wild-card.
B. Discovering Frequent Approximate Sequences
At the beginning select randomly one subsequence of
length W from each input sequence. These subsequences will
form our initial set of occurrences. Denote oi occurrence in
sequence i.Iteration step.
Step 1: Pick randomly one sequence i.
Step 2: Compute position weight matrix based on all
occurrences except oi Denote this position weight matrix P.
Step 3: Take each subsequence of sequence i of length W, and
compute a score of this subsequence according to matrix P.
Step 4: Choose new occurrence o′i randomly among all
subsequences of i of length W using probability distribution
defined by the score (higher score means higher probability).
Step 5: Replace oi with o′i in the set of occurrences.

98
ISSN: 2454-1532
Step 6: Repeat iteration steps, until some stopping condition is
met.
In model (L, d, k), L is the length of the pattern, d is the
maximum distance with in which two strings are considered
similar and k is the minimum support or frequency required
for a valid patterns. In proposed method we are considering
only those strings of length L that actually occur in the dataset
and compute the support for each of them by scanning the
dataset. This approach will not discover patterns as the model
string might not actually occur in the dataset even once.
C. LDK Algorithm Methodology
We first discuss our approach for Problem of
complex pattern matching where we are given a pattern P and
a text T and T can contain sets of characters. The basic idea of
our approach is as follows. We consider each position in T and
P as a class or a set for P each position is a set of one character
( i.e. a singleton set). For each character in the alphabet, we
construct a smaller similar problem, namely a restricted don’t
care matching problem, to be defined shortly, and solve them
separately. The results are then combined to get the final
result. Let us first formally define the restricted don’t care
matching problem.
Algorithm 1: LDK Algorithm for Complex Pattern Matching
Input:
1. Input sequence is composed of symbols from a discrete
alphabet sets.
Set of items/alphabets I= {a1, a2, . . . ., am}
Input Sequential database D = { x | ∀x ϵ I}
E.g. I = {A, B, C, D}
D = ACABBDACCAB
2. Values of L, d and k.
Where
L: Length of substring
D: Distance/number of mismatches allowed
K: Minimum Support
Output:
Set of (L, d, k) model Strings.
Op = {s1, s2, . . . ., sn}
Where
Si = {si1, si2, . . .,sim} is set of instances for si.
Op = {si| d(si, sij) ≤ d, |sij|=L and m ≥ k}
Algorithm 2: Frequent Pattern Matching
Freq Pattern (m Tree, root, L, d, k)
1. process();
2. get model Tree();
3. For i=0 to i<m Tree. size()
4. compute Support();
5. End For
Algorithm 3: Suffix Tree Formation
process()
Construct count suffix tree on input dataset.
Algorithm 4: Generate Model Tree
Get model Tree()
Construct suffix tree of depth L on strings of length L.
That occurs in the count suffix tree.
Algorithm 5: Computing Support Value
Compute Support()
1. old Matches=model. parent. matches;
2. if(model.parent.id=1)
3. new matches=expand Matches()
4. else
5. for k=0 to k<old Matches. size()
6. new Matches. Add All(expand Matches())
7. End for
8. set Matches(new Matches)
9. for i=0 to model Tree. size()
10. x=model Tree. get(i)
11. if(distance()<=d)
12. model. matches. add(x)
13. model. Model Support +=x. support;
14. End for
15. End
Frequently occurring pattern of model (L, d, k) a string of
length L that occurs minimum k times in the input dataset,
with each occurrence being within a Hamming distance of d
from the model string. Hamming distance between strings S
and P .Now we are ready to formally state the algorithm in the
form of Algorithm 1. The analysis of Algorithm 1 is
straightforward. Step 2 and Step 3 can be done implicitly,
while performing Step 4. Therefore in Step 1 we basically
perform Step 4 |Σ| times requiring O(|Σ|nlogm) time. Step 6
can be done incrementally in the loop and hence the total
running time would remain the same i.e. O(|Σ|n log m). It is
easy to verify that, if P contains sets of characters instead of T,
we just need to swap the definitions of T′ and P′ in Step 2 and
Step 3 respectively. We would also like to note that, if both T
and P are degenerate the algorithm wouldn’t work properly. In
the rest of this section this thesis practical efficiency
implementation technique of our approach.
IV. IMPLEMENTATION AND RESULTS
In this section we introduce two examples were usage of
pattern finding tools helped researchers to make new
discoveries. There are of course many such examples. The two
examples chosen here illustrate, types of discoveries that can
be made and how these discoveries can be verified by
biological experiment.
A. Performance Evaluation Using EBOLA Dataset
Proteins secreted by Mycobacterium EBOLA, the
causative agent of EBOLA, are often targets of immune
responses by an infected host that can lead to protection from
the disease. Thus the identification and study of these
secretory proteins is a first step to the design of more effective
vaccines targeted against this dreadful pathogen. Proteins are
directed to the outer membrane of bacterial cells by a short N-
terminal sequence of amino acids called a signal peptide.

99
ISSN: 2454-1532
The signal peptide, which remains attached to the internal
face of the membrane, is cleaved to release the protein to the
external environment. Conserved features of both the signal
peptide and cleavage site, such as length and amino acid
composition, make them amenable to identification by
sequence analysis.
The recently released nucleotide sequence of the M.
EBOLA genome was screened for these conserved features
and a number of positive secretory proteins were detected
[Gomez et al., 2000].Evidence suggests the existence of other
as yet undefined mechanisms for protein secretion in M.
EBOLA .Identifying a natural reservoir for Ebola virus has
eluded researchers for decades. Dataset for DNA This data
string is used as text data contains the information (Homo
Sapiens ,DNA ID, DNA Family.
B. Result Analysis Comparison Character
The below DNA sequence dataset has been taken for the
testing of the proposed algorithm. The DNA biological
sequence.
S∈Σ*of size n = 1024 and pattern P∈Σ*. Let S be the
following DNA sequence
The DNA Data for sequence S is very large. For different
patterns P’s the number of occurrences and the number of
comparisons is huge. To check whether the given pattern
presents in the sequence or not we need an efficient algorithm
with less comparison time and complexity. By the current
technique different patterns are analyzed and using these
results plots the graph. From the below experimental results,
improvement has be seen that LDK algorithm gives good
performance compared to the some of the popular methods.
reduces the time complexity during worst case from O((n-
m+1)m) to O(nm+1). This time complexity is hugely depends
on the selected prime number, q. So selecting the right prime
number gives this algorithm a satisfiable optimization in terms
of worst-case time complexity.
 Different pattern sizes has been taken from the DNA
sequence ranging from 1 to 20 randomly and tested by
1024 DNA characters. CPC ratio decreases and is less
than 1 in case of the proposed technique where as in other
cases CPC is more than 1 in some of the existing methods
Hierarchical Verification Algorithm (HVA) and Matching
with Depth First Search (MDFS).
 We can see the following fields for comparison of
different algorithms like pattern text, number of
characters in the pattern, number of occurrences of a
pattern, the proposed method and the number of
comparisons with comparisons per character.
 The number of comparisons per character (CPC ratio)
which is equal to (Number of comparisons/file size) can
be used as a measurement factor, this factor affects the
complexity time, and when it is decreased the complicity
also decreases. Reduction in number of comparisons.
 The ratio of comparisons per character has gradually
reduced and is less than 1.
 Suitable for unlimited size of the input file.
 Once the indexes are created for input sequence we need
not create them again.
 For each pattern we start our algorithm from the matching
character of the pattern which decreases the unnecessary
comparisons of other characters.
 It gives good performance for DNA related sequence
applications.
TABLE I. Comparison of CPC between algorithms.
Algorithm Comparisons Per Character (CPC)
HVS 4
MDFS 3
LDK 1
Fig. 2. Comparison of CPC between algorithms.
C. Result Analysis - Time Complexity
In Best case doesn’t differ much from the original
algorithm, but the in average case complexity can be improved
significantly. Due to imposing of constraint of matching
complex patterns the HVS algorithm and MDFS algorithms
takes more time than the LDK algorithm as shown on figure.
Fig. 3. Shows the results of time complexity.
V. CONCLUSION
Pattern discovery is an important area of bioinformatics.
The algorithms for pattern discovery use wide range of
computer science techniques, ranging from exhaustive search,

100
ISSN: 2454-1532
elaborate pruning techniques, efficient data structures, to
machine learning methods and iterative heuristics. The tools
developed by computer scientists are today commonly used in
many biological laboratories. They are important to handle
large scale data, for example in annotation of newly sequenced
genomes, and organization of proteins into families of related
sequences. They are also important in smaller scale thesis,
because they can be used to detect possible sites of interest
and assign putative structure or function to proteins. Thus they
can be used to guide biological experiments, decreasing the
time and money spent in discovering new biological
knowledge about the Ebola virus effecting of the peoples.
REFERENCES
[1] R. Agrawal and R. Srikant, ―Mining sequential patterns,‖ In 11th
International Conference on Data Engineering, 1995.
[2] R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A. I. Verkamo,
―Fast discovery of association rules,‖ 1996.
[3] R. Agrawal, T. Imielinsk, and A. Swami, ―Mining association rules
between sets of items in large databases,‖ In Proceedings of the
ACMSIGMOD Conference on Management of Data, Washington, D.C. ,
pp. 207-216, 1993.
[4] P. Hoffman, ―DNA visual and analytic data mining,‖ Proceedings of the
8th IEEE Visualization '97 Conference, 1997.
[5] P. Papapetrou, ―Discovering frequent poly-regions in DNA sequences,‖
Sixth IEEE International Conference on Data Mining-Workshops
(ICDMW'06), 2006.
[6] J. Hu, ―Mining sequence features for DNA-binding site prediction,‖
IEEE Symposium on Computational Intelligence in Bioinformatics and
Computational Biology, 2008. CIBCB '08, pp. 219-222, 2008.
[7] K.-S. Leung, ―Data mining on DNA sequences of hepatitis B virus,‖
IEEE/ACM transactions on computational biology and bioinformatics,
vol. 8, no. 2, 2011.
[8] B. Ding, ―Efficient mining of closed repetitive gapped subsequences
from a sequence database,‖ IEEE International Conference on Data
Engineering, pp. 1024 – 1035, 2009.
[9] S. Bai, ―The maximal frequent pattern mining of DNA sequence,‖ IEEE
International Conference on Granular Computing, pp. 23–26, 2009.
[10] A. Floratou, ―Efficient and accurate discovery of patterns in sequence
datasets,‖ IEEE 26th International Conference on Data Engineering
(ICDE), pp. 461–472, 2010.
[11] S. Li, ―An optimized algorithm for finding approximate tandem repeats
in DNA sequences,‖ IEEE Second International Workshop on Education
Technology and Computer Science, pp. 68-71, 2010
[12] C. C. Wang, ―Predicting DNA-binding locations and orientation on
proteins using knowledge-based learning of geometric properties,‖ IEEE
International Conference on Bioinformatics and Biomedicine, pp. 3–8,
2010.
[13] A. Floratou, ―Efficient and accurate discovery of patterns in sequence
data sets,‖ IEEE Transactions on Knowledge and Data Engineering, vol.
23, no. 8, pp. 1154-1168, 2011.
[14] K. Atteson, ―The exact probability of language-like patterns in
biomolecular sequences,‖ In Proceedings of the 6th
International
Conference on Intelligent Systems for Molecular Biology (ISMB), pp.
17–24, 1998.
[15] T. L. Bailey and C. Elkan, ―Fitting a mixture model by expectation
maximization to discover motifs in biopolymers,‖ In Proceedings of the
2nd International Conference on Intelligent Systems for Molecular
Biology (ISMB), pp. 28-36, 1994.
[16] T. L. Bailey and C. Elkan, ―Unsupervised learning of multiple motifs in
biopolymers using expectation maximization,‖ Machine Learning,
21(1/2), pp. 51–80, 1995.
[17] W. C. Barker, F. Pfeiffer, and D. G. George, ―Superfamily classification
in PIR-international protein sequence database,‖ Methods in
Enzymology, vol. 266, pp. 59-71, 1996.
[18] G. Benson, ―Tandem repeats finder: a program to analyze DNA
sequences,‖ Nucleic Acids Research, vol. 27, issue 2, pp. 573–580,
1999.
[19] M. Blanchette, B., Schwikowski, and M. Tompa, ―An exact algorithm to
identify motifs in orthologous sequences from multiple species,‖ In
Proceedings of the 8th
International Conference on Intelligent Systems
for Molecular Biology (ISMB), pp. 37–45, 2000.
[20] E. B. Bauer, E. Rivals, and M. Vingron, ―Computational approaches to
identify leucine zippers,‖ Nucleic Acids Research, vol. 26, issue 11, pp.
2740–2746, 1998.
[21] A. Brazma, I. Jonassen, I. Eidhammer, and D. Gilbert, ―Approaches to
the automatic discovery of patterns in biosequences,‖ Journal of
Computational Biology, vol. 5, issue 2, pp. 279–305, 1998.
[22] A. Califano, ―SPLASH: structural pattern localization analysis by
sequential histograms. Bioinformatics, vol. 16, issue 4, pp. 341–347,
2000.
[23] R. Cowan, ―Expected frequencies of DNA patterns using Whittle’s
formula,‖ Journal of Applied Probability, 28(4), pp. 886–892, 1991.
[24] E. Coward, and F. Drab- los, ―Detecting periodic patterns in biological
sequences,‖ Bioinformatics, vol. 14, issue 6, pp. 498–507, 1998.
[25] B. Dorohonceanu and C. G. N. Manning, ―Accelerating protein
classification using suffix trees,‖ In Proceedings of the 8th Inter-
national Conference on Intelligent Systems for Molecular Biology
(ISMB), pp. 128–133, 2000.
[26] R. Durbin, S. R. Eddy, A. Krogh, and G. Mitchison, ―Biological
Sequence Analysis,‖ Cambridge University Press, Calculating 30, 1998.

Efficiency of Using Sequence Discovery for Polymorphism in DNA Sequence

Recommended

More Related Content

What's hot (18)

Viewers also liked (15)

Similar to Efficiency of Using Sequence Discovery for Polymorphism in DNA Sequence (20)

More from IJSTA (20)

Recently uploaded (20)

Efficiency of Using Sequence Discovery for Polymorphism in DNA Sequence