See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/347132039

Bioinformatics Database Query Performance and Optimization

Article · September 2020

DOI: 10.35940/ijrte.C4666.099320

CITATIONS READS

0 58

3 authors, including:

Edy Budiman Ummul Hairah

Universitas Mulawarman Universitas Mulawarman
82 PUBLICATIONS   308 CITATIONS    19 PUBLICATIONS   72 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Integrated multi criteria decision making for a destitute problem View project

prediction View project

All content following this page was uploaded by Edy Budiman on 15 April 2021.

The user has requested enhancement of the downloaded file.

 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-9 Issue-3, September 2020

Bioinformatics Database Query Performance and

Optimization
Edy Budiman, Andi Tejawati, Ummul Hairah

 The query optimization method we propose is a file

Abstract: Borneo bioinformatics portal test is a critical element resource efficiency issue or source-code used in the
of SQA and represents a comprehensive review of specifications, development of bioinformatics webportal.
design and coding. The test represents an abnormality in the The bioinformatics portal testing objects was portal was
development of the portal. A series of tests systematically reveals
several different types of errors. This study aims to evaluate the developed by the author to manage data on plant biodiversity
performance and optimization of Borneo's Bioinformatics portal on the Borneo island[16]. The portal is built using the
with a series test activities using the Web Performance Model-View-Controller (MVC) concept based on Laravel's
Optimization methodology. Testing query performance with PHP framework and utilizes the Eloquent ORM features in
measuring the response time and page loading timings from the query relations data mapping[17]. Now, the Borneo's
object relationship mapping (ORM) model Laravel PHP
framework in offline and online. For optimization, we set a
Bioinformatics Portal currently contains 233 record
pre-test and post-test scenario to evaluate the efficiency medicinal plant, 1482 record tree species and 86 record wood
performance test results. The results study found that the query species include 80 record bamboo species[18].
relation model, parsing script (javaScript and CSS), service scale This study aims to evaluate with performance testing of the
and dimension images in the interaction process to the database query model and optimizes the script (source code) on the
are the dominant resources affecting the performance of the
Borneo Bioinformatics portal. For query performance testing
Bioinformatics portal. Performance optimization through
determining the appropriate query relation model, minify and we measure the response time of the query relations the
defer parsing script or combine images using CSS sprites to Laravel eloquent ORM model and page load Timings of
reduce scala image. Borneo Bioinformatics portal. Whereas for script
Keywords : Bioinformatics, query, database relationship, optimization, we measure the efficiency of source code usage;
ORM. PHP, JavaScript and CSS using the pretest and posttest
scenarios in Borneo's Bioinformatics portal.
I. INTRODUCTION Research contribution: A common problem of the
Bioinformatics portal is web performance optimization. This
One of the optimization parameters to improve the study result is a proposal for a query optimization method on
the Bioinformatics web portal which is built within the PHP
performance of a bioinformatics resource portal is the
Laravel framework, through an efficiency optimization
database query model. In In overcoming the challenges of
approach to source-code (scripts) or other data resource files.
complex queries, several methods of query RDMBS
modelling proposed in [1], under bioinformatics domains.
II. RELATED WORK
Various approaches can be applied to improve the
performance of bioinformatics portals, can uses perform
A. Literature
custom queries with a compilation of bioinformatics
A brief literature review of related work in the area of
scenarios[2], database integrated system based on SOAP web
Bioinformatics portal and database performance evaluation
service[3] and extensive web services[4], resource portal[5],
resulted in the identification of several studies, such as;
temporal database[6] or via caching[7], DOM interactive and
Bultet, L. et al, the SIB Swiss Institute of Bioinformatics’
loaded time content[8], the graph-based database partitioning
resources: focus on curated databases[19], Bioinformatics
method (parallel query)[9], efficiency of the Biodiversity
web portals[20], ExPASy: SIB bioinformatics resource
resource portal with of the key performance indicators[10],
portal[21], Ethnobotany database: Exploring diversity
query optimization through the Grid data sources
medicinal plants of Dayak Tribe Borneo[22], Manycore
architecture[11], taxonomy of plants Nomenclature[12],
High-Performance Computing in Bioinformatics[23], A
NoSQL Biological Databases[13], with ontology and
comparison of microbial genome web portals[24], BioPortal:
taxonomy model[14] and Ontologies for Query
A Portal for Deployment of Bioinformatics Applications on
Relaxation[15], etc.
Cluster and Grid Environments[25], mobile internet services
Revised Manuscript Received on September 25, 2020.
performance in borneo[26], etc. Other than that Sukhpuneet
* Correspondence Author Kaur et. al An Empirical Performance Evaluation of
Edy Budiman*, Informatics Department, Universitas Mulawarman, Universities Website International Journal of Computer
Samarinda, Indonesia. Email: edy.budiman@fkti.unmul.ac.id Applications[27], and D.
Andi Tejawati, Informatics Department, Universitas Mulawarman,
Samarinda, Indonesia. Email: andi.tejawati@gmail.com
Ummul Hairah, Informatics Department, Universitas Mulawarman,
Samarinda, Indonesia. Email: ummihairah@gmail.com

Published By:
Retrieval Number: 100.1/ijrte.C4666099320 Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.C4666.099320 581 and Sciences Publication
 Bioinformatics Database Query Performance and Optimization

Luna in Government web portals performance evaluation (F) >=26 CSS >= 11
>= 5 file size >= 6 unminified
using data envelopment analysis[28], where according to 0 <= S
< 500b
or inline
components
duplicated
< 50 Style JS or CSS
these studies the quality assurance of a website depends on
automated testing tools which lower costs and increase its Matrix table-keys according to[32]:
efficiency. The performance of a website can be an important  Rule: The YSlow performance rule
factor for its success. It depends on the main speed factor. If  Weight. How this performance rule is weighted in the
the website speed is fast then the performance automatically overall page analyis grade
increases.  Points. Number of points deducted per offender
(performance infraction occurance), from a total of
B. Metric for Perfomance
100 per rule
Performance can be evaluated using tools that break down  Score Computation. The formula used to compute the
the resources and components on the bioinformatics web site. final score per rule
There are a wide variety of automated site testing tools  Grades from A to F. How many components/offenders is
available. necessary to reach grades from A to F.
 Google PageSpeed Insights (PSI), Refers to [29]
PageSpeed Insights (PSI) reports on the performance of a C. Borneo Bioinformatics Portal
page on both mobile and desktop devices and provides One of the efforts to manage biodiversity is through data
suggestions on how that page may be improved. PSI provides and information support. Data and information on
a score which summarizes the page’s performance. This score biodiversity need to be continued efforts are made to be
is determined by running Lighthouse to collect and analyze added, both in species diversity, habitat, population, and
lab data about the page. A score of 90 or above is considered distribution. Records of 47,910 species of Indonesian
good. 50 to 90 is a score that needs improvement, and below biodiversity [33] are estimated still far less than the potential
50 is considered poor. PSI also classifies field data into 3 that actually exists. It is necessary to increase the intensity of
buckets, describing experiences deemed good, needs the implementation of identification and inventory of
improvement, or poor (see Tabel-I). biodiversity in the field, and on the other hand, a database
Tabel- I: Classifying Good, Needs Improvement, Poor[29] system that is able to collect data and information that is
Good Needs Improvement Poor spread across various circles is needed
FCP [0, (1000ms, 3000ms] over 3000ms Borneo bioinformatics portal as a data management system
1000ms]
FID [0, 100ms] (100ms, 300ms] over 300ms and information on endemic plants for the island of Borneo,
LCP [0, (2500ms, 4000ms] over 4000ms Kalimantan, Indonesia. Borneo Bioinformatics is an example
2500ms] of an information system which presents taxonomic data,
CLS [0, 0.1] (0.1, 0.25] over 0.25
form an ontology model which can serve, mapping data as
 Yslow from Yahoo, Refers to [30] YSlow grades web page information regarding data descriptions and relationships
based on one of three predefined rulesets or a user-defined between taxons in accordance with taxonomic levels based on
ruleset. It offers suggestions for improving the page's data stored in the database[34].
performance, summarizes the page's components, displays The portal online is accessed in url
statistics about the page, and provides tools for performance https://www.borneodiversity.org/index. The BBIS interface
analysis [31]. YSlow's web page analysis is based on the 23 of is shown in Figure 1.
these 34 rules that are testable [30] (see Table-II).
Tabel -II. Rule weights of YSlow V2 Ruleset [32].
Remove
Compress
Avoid CSS Minify JScript duplicate
Rule component
expressions and CSS JScript and
with GZip
CSS

0 to 5
(A) expressions 0 or 1 0 to 2
0 file size <
90 <= S on CSS or unminified duplicated
500b
<= 100 inline component JS or CSS
STYLE

6 to 10
(B) 1 file size < 3 or 4
expressions 2 unminified
80 <= S 500b of any duplicated
on CSS or components
< 90 type
inline Style
JS or CSS Fig. 1. Screenshot of the borneo bioinformatics portal
The first stage of data collection, to date, the system have
11 to 15
(C)
2 file size < expressions 3 unminified
5 or 6 recorded 233 Medicinal data, 1482 tree species, 86 types of
70 <= S duplicated
< 80
500b on CSS or components
JS or CSS wood, and 80 types of bamboo. Until now, the data collection
inline Style
process is still continuing.
(D) 3 file size < 16 to 20 7 or 8
4 unminified
60 <= S 500b of any CSS or duplicated
components
< 70 type inline Style JS or CSS

4 21 to 25
(E) 9 or 10
uncompress expressions 5 unminified
50 <= S duplicated
ed or file on CSS or components
< 60 JS or CSS
size < 500b inline Style

Published By:
Retrieval Number: 100.1/ijrte.C4666099320 Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.C4666.099320 582 and Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-9 Issue-3, September 2020

III. METHODOLOGY Timings Connection duration

Backend duration
This study evaluates the performance of the query model Time to First Byte (TTFB)
and optimizes the script on the Borneo Bioinformatics portal. DOM interactive-content loaded time
First paint time
For query performance testing, we measured the response First contentful paint time
times of the query Object Relationship Mapping (ORM) Onload time
model and page loads of the Borneo Bioinformatics portal. As
B. Equipment and Tools
for script optimization, we measure the efficiency of using
source code; PHP, JavaScript and CSS use the pretest and The performance testing and optimization tools used are
posttest scenarios on the Borneo Bioinformatics portal. shown in Table IV
Table- IV: equioment and tools
A. Performance Testing Methodologies Equipments Specifications
 For query bioinformatics portal: the Borneo Hardware: 2.5 GHz Dual Core Intel Core i5
Bioinformatics Portal is built based on Laravel PHP Laptop Platforms 8GB-4GB 2666MHz, 4 GB DDR4-Memory
Server Server version: 10.1.31-MariaDB -
framework (“active record” pattern), We test query MariaDB Serve
performance utilizing the Laravel Object Relations Mapping Software:
model features available in interactions to the database. Laravel ORM Package Laravel debugbar versi 5.5
Apache Xampp PHP version 7.1
Bioinformatics portal url: http://borneodiversity.org
GTMetrix Tools PSI Google and Yslow yahoo

C. Performance Test Scenarios

Performance
Optimization Performance test and optimization scenarios (Query -
Script) used are seen in Table V
JavaScript (JS) Table- V: Performance test scenarios
Bioinformatics Code (Scripts)
Cascading Style Scenarios Descriptions
Sheets (CSS)
Query performance: Using Eloquent - Query Builder (ORM vs Non
Result Call
(view) (route) Image ORM) to data search process and view in
Bioinformatics portal.
Measurement data: Offline and Online
Object Relationship Model a. One-to-one a. Relation data wood to medicinal
(ORM)
b. One-to-many b. Relation data tree to wood
Generate
SQL Mysql c. Many-to-many c. Relation data tree to wood
One to One Many to Many
Query d. Has-Many-Through d. Relation data medicinal to data tree and
One to Many Has Many Through wood
Result

DB Connection
tree medicinal wood taxonomy
Fig. 2. Performance test design id id
id id
The test is to get the query execution time (response) based tree kingdom
leaf_flower wood
on the relations formed. Two types of testing are performed, latin division
rod_root botany
synonym class
offline and online for each relationship ie; one to one, one to fruit_seed local
ordo
local chemical synonym
many, many to many, and has many through using the Laravel image family
information_ habitus
Debugbar package. An overview of the performance test ecology research picture
genius
design is seen in Figure 2. endemic species
efficacy medicinal_id
sub species
 For optimization performance bioinformatics portal: high descriptions descriptions
stem_color varietas
performance testing and optimization using the pretest and
sap_color
post-test scenarios. Pretest is the performance of the portal information_
before optimization, Posttest is the performance portal after research
optimization. Performance measurement parameters are descriptions
Script(Javascript and CSS), images and page load timings,
which are files that affect the performance of a Bioinformatics Fig. 3. Test tables for query performance (data relations)
portal. Table 1 is the parameter as a performance test metric. Figure 3 for testing the relation scheme One To One
Table- III: Performance optimization metrics relation (hasOne), between medicinal table and wood table,
Parameter Metric scheme One to Many relation (hasMany) between tree table
Serve scaled images and wood table, scheme Many to Many relation
Image Optimize images
Image dimensions
(belogsToMany) between tree table and wood table, and for
Minify CSS and Javascript scheme Has Many Through relation between medicinal table
Defer parsing JavaScript to tree table and wood table.
Script Inline small CSS and JavaScript
(CSS and Combine images using CSS sprites
JavaScript Avoid CSS @import
Duplicate JavaScript and CSS
Avoid CSS expressions
Page Load Redirect duration

Published By:
Retrieval Number: 100.1/ijrte.C4666099320 Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.C4666.099320 583 and Sciences Publication
 Bioinformatics Database Query Performance and Optimization

IV. RESULTS AND DISCUSSION relationship between tree table and wood table. The results of
One To Many relations query test are shown in Table VI and
A. Result: Query Performance Figure 5.
The query test gets the execution time to display data based Table -VI: Query performance for One To Many relation
on the relations schema in the scenario. Tests are carried out Online Offline
on each relationship, ie, one to one, one to many, many to Test ORM Non ORM ORM Non ORM
many, and has many through, and the offline and online 1 22930 115.6 372.59 68.02
testing.
2 20160 260.96 383.09 15.83
The query performance test results are as follows:
 Result: One-To-One relation 3 22400 328.45 383.28 66.14
This test is done to get the time from the results of the query 4 22400 268.56 423.1 14.51
execution for the One to One relation by using Eloquent ORM 5 21420 252.57 380.3 14.84
and Query Builder to display data. One-to-One relationship
6 33460 161.03 381.34 5.94
between wood table and medicinal table. The results of One
7 42700 367.03 387.52 6.1
To One relations query test are shown in Table V:
Table- V: Query performance for One To One relation 8 7700 231.46 379.34 6.76
Online Offline 9 29420 411.13 402.37 6.58
Test ORM Non ORM ORM Non ORM 10 25960 456.15 401.4 6.3
1 9710 58.63 143.31 7.25 389.43
Avg. 24855 285.294 21.102
3
2 6780 250.75 137.02 6.26
3 5680 324.65 148.93 6.77
4 7870 40.65 148.17 7.74
5 3550 124.81 136.12 7.27
6 4920 119.59 135.75 7.06
7 5470 140.65 137.82 7.69
8 2190 447 144.76 7.64
9 2500 574.75 137.11 7.17
10 2750 49.19 137.67 6.2
140.66
Avg. 5142 213.067 7.105
6

Fig. 5. Performance query One to Many relationship

From the test in Figure 5, it is known that the One To Many
relation with the ORM response time of average for online is
24855 ms to execute the field data and without ORM is
285.294 ms.
 Result: Many-To-Many relation
This test is done to get the time from the results of the query
execution for the Many to Many relation by using Eloquent
ORM and Query Builder to display data. One to Many
relationship between tree table and wood table. The results of
Many To Many relations query test are shown in Figure 6 and
Table VII.
Table- VII: Query test for Many To Many relation
Online Offline
Fig. 4. Performance query ORM One to One relationship
Test ORM Non ORM ORM Non ORM
The results of testing the ORMe One To One relationship
query in Figure 4, explain the average results of the 1 383.74 257.24 13.77 7.61
comparison performance between the use of ORM online and 2 740.64 157.46 14.22 5.2
offline, the difference in response time average is 5595.54 ms 3 349.78 179.4 12.45 5.47
for online, and offline of 140.66 ms query execution time,
4 411.19 259.61 12.56 4.81
with 2 table relations showing all the data fields there is. It
5 363 288.02 11.93 4.72
was found that ORM offline it is faster.
 Result: One-To-Many relation 6 799.72 341.32 13.27 6.79
This test is done to get the time from the results of the query 7 228.31 318.83 12.42 5.19
execution for the One to Many relation by using Eloquent
ORM and Query Builder to display data. One to Many

Published By:
Retrieval Number: 100.1/ijrte.C4666099320 Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.C4666.099320 584 and Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-9 Issue-3, September 2020

8 429.79 407.85 12.21 5.16 average 3189.65 ms and with an average time for Non ORM
9 146.54 232.95 11.28 6.74 of 58.008 ms to execute the field data.
To get a comparison between offline and online testing
10 432.81 122.81 11.91 4.97
Then a summary of the average response time of each relation
Avg. 428.552 256.549 12.602 5.667 as seen in Table IX.
Table -IX: Summary offline-online query test relation
Online Offline
Relationshi Non
ORM Non ORM ORM
p ORM
One to One 5142 213.067 140.666 7.105
One to
24855 285.294 389.433 21.102
Many
Many To
428.552 256.549 12.602 5.666
Many
Has Many
3189.65 58.088 84.948 5.731
Through
156.912
8403.801 203.2495 9.901
3
When testing an impedence mismatch case, it occurs when
Fig. 6. Performance query Many to Many relationship
there is a mapping problem in the database relation in
From the test in Figure 6 and Table VII, it is known that the
displaying details of plant data that displays data from
Many To Many relation without ORM response time offline
columns with the same name even though the columns are in
average is 5.667 ms and online average time is 256.549 ms to
different tables.
execute the field data.
 Result: Has Many Through relation B. Results: Borneo Bioinformatics Optimization
This test is done to get the time from the results of the query  PreTest
execution for the Has Many Through relation by using The results of the Preliminary Test (PreTest) on the
Eloquent ORM and Query Builder to display data. One to Borneo Bioinformatics portal are presented in Figure 8.
Many relationship between tree table and wood table. The
results of Has Many Through relations query test are shown in
Table VIII and Figure 7.

Fig. 8. Screenshot Prestest performance of the portal

Fig. 7. Performance query One to Many relationship The results of the pretest performance reduction obtained a
Table- VIII: Query test for Has Many Through relation Grade F score (48%) for PageSpeed, and a YSlow's Score
with Grade F (48%). For page details (Full Load Time is 28.1
Online Offline
seconds, with a total page size of 9.34MB and 311 Requests).
Test ORM Non ORM ORM Non ORM
Figure 9 (pretest) presents a list of recommendations from
1 5440 29.33 84.25 5.56 PageSpeed with the performance scores obtained in the
2 3174.2 70.05 80.87 6.44 Pre-Test. There are seven (7) recommended items that score
3 4392.3 98.72 85.95 5.39 very low, i.e. "Serve scaled images, Defer parsing of
JavaScript, Optimize images, Inline small CSS, Minify
4 3580 89.52 83.16 5.97
JavaScript” each obtain with Grade F (0). For “Inline small
5 2920 24.51 84.99 5.54
JavaScript" with Grade F(41), Avoid CSS@import with
6 3970 85.56 88.48 4.95 Grade C(73).
7 2650 28.73 86.75 5.62
8 1490 35.65 83.97 6.52
9 1640 24.83 85.56 5.72
10 2640 93.98 85.5 5.600
Avg. 3189.65 58.088 84.948 5.731

From the test in Figure 7, it is known that the Has Many

Through relation with the ORM response time online is

Published By:
Retrieval Number: 100.1/ijrte.C4666099320 Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.C4666.099320 585 and Sciences Publication
 Bioinformatics Database Query Performance and Optimization

CSS, Remove duplicate JavaScript and CSS and Make

JavaScript and CSS eksternal” each obtain Grade F(0). For
other recommended items, generally get Grade A with a score
of 90 – 100.
 PostTest
The Post-test given to the portal after completing the
improvements based on the recommendations of the pretest
results (Tabel X), is to measure the performance achievement
and optimization of the website.

Fig. 9. Screenshoot pretest from google pagespeed

For other recommended items, generally get Grade A with
a score of 91 - 100. The low scoring recommendations and
optimization solutions from the PageSpeed pre-test results are
presented in Table X.
Table X. Performance values of pagespeed
recommendations for optimization
Grade
Recommendation Type Optimization
Score
Minimized serving scaled
Serve scaled
F (0) Images images can save 3.3MiB (77%
images
reduction)

Defer parsing of Defer parsing JS to reduce

F (0) JS
JavaScript blocking of page rendering

Optimize images to reduce

Optimize images F (0) Images
their size by 965.0KiB (12%)

Inlining the response in

Inline small CSS F (0) CSS
HTML can reduce-rendering.

Minify JavaScript for the

Minify JavaScript F (0) JS
resource by 31% reduction)

Strip unnecessary characters

Minify CSS F (0) CSS from JavaScript and CSS to
speed up download times

Inlining the response in

Inline small
F(41) Js HTML can reduce blocking of
JavaScript
page rendering

Avoid The external stylesheets were Fig. 11. Screenshot Posttest PageSpeed performance
C(73) CSS
CSS@import included in using @import The performance score of the main portal page after
Whereas for the pretest recommendation from YSlow Post-Test gets Grade C (77%) for Pagespeed
yahoo is presented in Figure 10. recommendation, and for YSlow with Grade B(80%).
Regarding Detail Pages' performance to get score for Full
Load Time is 3.5 seconds, Total Page Size is 9.00MB of 317
Requests. The results of the PostTest measurement in Figure
11 an 12 have shown that the score of each recommendation
item has been optimized which shows a good value. For
"Serve scaled images with Grade C (75)," Defer parsing of
JavaScript with Grade A (99) and "Minify JavaScript with
Grade A (99), Optimize images with Grade A(97), Inline small
CSS and JavaScript with Grade A(100), Minify CSS with
Grade B(83) and JavaScript with Grade A(99). For Avoid
CSS@import with Grade A (100).
Fig. 10: Screenshot prestest YSlow performance
Figure 10 (prestest) presents a list of recommendations
from YSlow with the performance scores obtained in the
PreTest. There are four (4) recommended items that score
very low, i.e. " Avoid CSS expressions, Minify JavaScript and

Published By:
Retrieval Number: 100.1/ijrte.C4666099320 Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.C4666.099320 586 and Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-9 Issue-3, September 2020

a higher response time than without ORM. however, relations

without ORM cause impendace mismatch problems in the
data mapping process.
 The Efficiency Performance of the Bioinformatics Portal
based on the index generated from the PageSpeed Google and
YSlow recommendations from Yahoo is still low, and of
course it will directly affect the portal's performance.
The results of the evaluation of the bioinformatic portal
work in the future require performance improvements and
optimization, adjustments to the query relation model, and the
appropriate coding structure and serve scale image.

Fig. 12. Screenshot Posttest YSlow performance V. CONCLUSION

Furthermore, the measurement of the Pretest vs PostTest Indonesia is one of the countries with the highest level of
loading times is presented in the Figure 13. threat to biodiversity [35]. To inhibit the rate of extinction of
biodiversity that is accelerated exponentially by human
activities, efforts to conserve biodiversity need to be
optimized, both in its natural habitat (in-situ conservation)
and artificially outside its habitat (external conservation).
Borneo BioInformatics as large-scale system consists of
datafields, images and relations, and the data continues to
increase and grow. The portal issue is performance
optimization. respon time needed to access web pages
(page-loading) is still high, becomes slow when accessed,
takes up hosting server resources, and other problems.
The results of the study found that the query relation model,
Fig. 13. Screenshot pretest-posttest Page loading timings parsing script (javaScript and CSS), service scale and
In principle, the page loading speed of a website will dimension images in the interaction process to the database
improve the user experience in accessing information. In are the dominant resources affecting the performance of the
principle, the page loading speed of a website will improve Bioinformatics portal. Performance optimization through
the user experience in accessing information. The determining the appropriate query relation model, minify and
Bioinformatics portals that reference external objects most of defer parsing script or combine images using CSS sprites to
the page load time is spent on HTTP requests separate from reduce scala image.
images, javascript, and stylesheets. In particular, you can feel
the effect of the HTTP request on internet connection and ACKNOWLEDGMENT
website page load time. This research was funded by the Higher Education
Institution Operational Assistance Fund (BOPTN), Dept. of
C. Discussion informatics, Faculty of engineering, Mulawarman University,
The bioinformatics portal test is a critical element of SQA Samarinda, east-Kalimantan province, Indonesia.
and represents a comprehensive review of specifications,
design and coding. The test represents an abnormality in the REFERENCES
development of the portal. A series of tests systematically 1. J. Y. Chen, J. V. Carlis, and N. Gao, “A complex biological databases
reveals several different types of errors. querying method,” Proceedings of the ACM Symposium on Applied
This study analyzes the performance of the Borneo Computing, vol. 1, no. March 2005, pp. 110–114, 2005, doi:
Bioinformatics portal with a series of tests using measuring 10.1145/1066677.1066708.
2. A. Messina, A. Fiannaca, L. La Paglia, M. La Rosa, and A. Urso,
and performance optimization tools. Based on the results of “BioGraph: A web application and a graph database for querying and
testing the response time for online access queries on four analyzing bioinformatics resources,” BMC Systems Biology, 2018, doi:
database relationship models, an average of 5142 ms is 10.1186/s12918-018-0616-4.
3. H. Haviluddin, E. Budiman, and N. F. Hidayat, “A database integrated
obtained for One to One, One to Many 24,855 ms, Many To system based on SOAP web service,” TEM Journal, 2019, doi:
Many 428,552 ms and Has Many Through an average of 10.18421/TEM83-12.
3189.65. In script testing, image and loading time of portal 4. A. Kalderimis et al., “InterMine: Extensive web services for modern
biology,” Nucleic Acids Research, 2014, doi: 10.1093/nar/gku301.
pages that apply the preTest-PostTest scenario, the 5. E. Budiman, N. Puspitasari, Haerullah, M. Jamil, M. Wati, and A.
optimization of the PageSpeed performance score is F (84%) Saudek, “Evaluation of the bioinformatics resource portal,” in
to C (74%), YSlow from pretest F (84%) to B (80%). Proceedings - 2018 3rd International Conference on Information
Technology, Information Systems and Electrical Engineering,
Likewise, the page load time, from 28.1 seconds to 3.5
ICITISEE 2018, 2018, pp. 54–59, doi:
seconds. 10.1109/ICITISEE.2018.8720973.
Based on the evaluation results, the following were found:
 The query relation model based on the ORM framework
makes it easier for database interactions (MVC), but time has

Published By:
Retrieval Number: 100.1/ijrte.C4666099320 Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.C4666.099320 587 and Sciences Publication
 Bioinformatics Database Query Performance and Optimization

6. J. Chomicki and D. Toman, “Temporal Databases,” in Foundations of 28. D. E. Luna, L. F. Luna-Reyes, J. R. Gil-Garcia, and R.
Artificial Intelligence, Volume 1., L. V. M. Fisher, D. Gabbay, Ed. Sandoval-Almazán, “Government web portals performance evaluation
Elsevier B.V., 2005, pp. 429–467. using data envelopment analysis,” 2011, doi:
7. P. Cybula, H. Kozankiewicz, K. Stencel, and K. Subieta, “Optimization 10.1145/2037556.2037617.
of distributed queries in grid via caching,” 2005, doi: 29. D. google, “About PageSpeed Insights,” developers.google.com.
10.1007/11575863_58. https://developers.google.com/speed/docs/insights/v5/.
8. E. Budiman, N. Puspitasari, S. N. Alam, T. M. A. Akbar, Haeruddin, 30. Marcelduran, “Web Performance Best Practices and Rules,” yslow.org.
and D. Indra, “Performance analysis of the resource loading time for http://yslow.org/.
borneo biodiversity information system,” 2018, doi: 31. Carbon60, “Recommendations,” gtmetrix.com.
10.1109/IAC.2018.8780515. https://gtmetrix.com/recommendations.html (accessed Jun. 06, 2020).
9. S. Wu, F. Li, S. Mehrotra, and B. C. Ooi, “Query optimization for 32. marcelduran, “YSlow Ruleset Matrix,” yslow.org.
massively parallel data processing,” 2011, doi: http://yslow.org/ruleset-matrix/.
10.1145/2038916.2038928. 33. Widjaja, Kekinian Keanekaragaman Hayati Indonesia 2014. 2014.
10. E. Budiman, N. Puspitasari, M. Wati, J. A. Widians, and Haviluddin, 34. E. Budiman, N. Puspitasari, M. Wati, Haviluddin, and R. Rahim,
“Web Performance Optimization Techniques for Biodiversity Resource “Model Framework for Development of Biodiversity Information
Portal,” Journal of Physics: Conference Series, vol. 1230, no. 1, 2019, Systems,” Journal of Physics: Conference Series, vol. 1230, no. 1,
doi: 10.1088/1742-6596/1230/1/012011. 2019, doi: 10.1088/1742-6596/1230/1/012012.
11. L. Zamboulis, N. Martin, and A. Poulovassilis, “Query performance 35. D. F. R. A. Cleary and L. DeVantier, “Indonesia: Threats to the
evaluation of an architecture for fine-grained integration of Country’s Biodiversity,” Encyclopedia of Environmental Health, no.
heterogeneous grid data sources,” Future Generation Computer November 2017, pp. 622–632, 2011, doi:
Systems, 2010, doi: 10.1016/j.future.2010.05.008. 10.1016/B978-0-444-52272-6.00504-3.
12. E. Budiman and S. N. Alam, “Database: Taxonomy of plants
Nomenclature for borneo biodiversity information system,” 2018, doi:
10.1109/IAC.2017.8280642.
AUTHORS PROFILE
13. L. Caroprese, E. Zumpano, and E. Vocaturo, “No SQL Database
Management Systems for Big Data,” International Journal of
Engineering and Advanced Technology, vol. 9, no. 5, pp. 21–26, 2020, Edy Budiman is member of the Association for
doi: 10.35940/ijeat.D9145.069520. Computing Machinery (ACM), member of Institute of
14. N. Puspitasari and E. Budiman, “Evaluation of Borneo’s Biodiversity Electrical and Electronics Engineers (IEEE), and member
Information System,” 2018 Electrical Power, Electronics, of APTIKOM (Asosiasi Pendidikan Tinggi Informatika
Communications, Controls and Informatics Seminar, EECCIS 2018, dan Komputer) and mem-ber of The Institution of
pp. 434–439, 2019, doi: 10.1109/EECCIS.2018.8692955. Engineers Indonesia (PII). Currently, he is actively
15. N. K. Gundla and Z. Chen, “Creating NoSQL Biological Databases with teaching and researching. As a writer on several journals and conferences, he
Ontologies for Query Relaxation,” 2016, doi: focuses his research on mobile network issues, performance and
10.1016/j.procs.2016.07.120. mobile-based apps.
16. N. Dengen, E. Budiman, J. A. Widians, M. Wati, U. Hairah, and M.
Ugiarto, “Biodiversity information system: Tropical rainforest borneo
and traditional knowledge ethnic of dayak,” Journal of
Telecommunication, Electronic and Computer Engineering, vol. 10, no.
1–9, 2018.
17. E. Budiman, M. Jamil, U. Hairah, H. Jati, and Rosmasari, “Eloquent
object relational mapping models for biodiversity information system,”
in 2017 4th International Conference on Computer Applications and
Information Processing Technology (CAIPT), Aug. 2017, vol.
2018-Janua, pp. 1–5, doi: 10.1109/CAIPT.2017.8320662.
18. U. Hairah, A. Tejawati, E. Budiman, and F. Agus, “Borneo biodiversity:
Exploring endemic tree species and wood characteristics,” in
Proceeding - 2017 3rd International Conference on Science in
Information Technology: Theory and Application of IT for Education,
Industry and Society in Big Data Era, ICSITech 2017, 2017, vol.
2018-Janua, pp. 435–440, doi: 10.1109/ICSITech.2017.8257152.
19. L. A. Bultet et al., “The SIB Swiss Institute of bioinformatics’ resources:
Focus on curated databases,” Nucleic Acids Research, vol. 44, no. D1,
pp. D27–D37, 2016, doi: 10.1093/nar/gkv1310.
20. M. Cannataro and P. Veltri, “Bioinformatics web portals,” in Selected
Readings on Database Technologies and Applications, 2008.
21. P. Artimo et al., “ExPASy: SIB bioinformatics resource portal,” Nucleic
Acids Research, 2012, doi: 10.1093/nar/gks400.
22. Haeruddin, H. Johan, U. Hairah, and E. Budiman, “Ethnobotany
database: Exploring diversity medicinal plants of Dayak Tribe Borneo,”
in International Conference on Electrical Engineering, Computer
Science and Informatics (EECSI), 2017, vol. 2017-Decem, doi:
10.1109/EECSI.2017.8239094.
23. J.-S. Varré, B. Schmidt, S. Janot, and M. Giraud, “Manycore
High-Performance Computing in Bioinformatics,” 2011.
24. P. D. Karp et al., “A comparison of microbial genome web portals,”
Frontiers in Microbiology. 2019, doi: 10.3389/fmicb.2019.00208.
25. W. W. Li et al., “Building cyberinfrastructure for bioinformatics using
service oriented architecture,” 2006, doi:
10.1109/ccgrid.2006.1630932.
26. E. Budiman and S. N. Alam, “User perceptions of mobile internet
services performance in borneo,” in 2017 Second International
Conference on Informatics and Computing (ICIC), Nov. 2017, vol.
2018-Janua, pp. 1–6, doi: 10.1109/IAC.2017.8280643.
27. S. Kaur, K. Kaur, and P. Kaur, “An Empirical Performance Evaluation
of Universities Website,” International Journal of Computer
Applications, 2016, doi: 10.5120/ijca2016910922.

Published By:
Retrieval Number: 100.1/ijrte.C4666099320 Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.C4666.099320
View publication stats
588 and Sciences Publication