Title - University of Pittsburgh

signtruculentBiotechnology

Oct 2, 2013 (3 years and 6 months ago)

68 views

Going Against Goliath


23
rd

May 2010


Katrina Kurtz
, MLIS

Carrie
Iwema
, PhD, MLS

Ansuman

Chattopadhyay
, PhD

Health Sciences Library System

University of Pittsburgh


http://www.hsls.pitt.edu/guides/genetics


Photo credit: http://www.uncabaret.com/node?page=5

Bioinformatics

Tools


Answer

biological

questions



Made by
informatics

researchers



Freely

available

to the public



Published

in
scholarly

journals



http://www.hsls.pitt.edu/guides/genetics

Bioinformatics

Tools


Examples
:


Protein

structure
prediction



Sequence

alignment



Databases

of
genes

associated

with

disease



Protein

interactions



http://www.hsls.pitt.edu/guides/genetics

Who

Uses
Bioinformatics

Tools?


Biologists:


Faculty researchers


Graduate students


Undergraduate students



Doctors who do biomedical research



Biomedical informatics researchers

http://www.hsls.pitt.edu/guides/genetics

Online
Bioinformatics

Resources

Collection


Manually

curated




Searchable

collection of
bioinformatics

tools



Freely

available

to the public



2696
resources



http://www.hsls.pitt.edu/guides/genetics/obrc/

http://www.hsls.pitt.edu/guides/genetics

http://www.hsls.pitt.edu/guides/genetics

http://www.hsls.pitt.edu/guides/genetics

http://www.hsls.pitt.edu/guides/genetics

Common
Criticisms




Why

bother
?




Everyone

uses Google.




Google
is

taking

over.

http://www.hsls.pitt.edu/guides/genetics

Is Google
Actually

Better
?

http://www.hsls.pitt.edu/guides/genetics


Large
amount

of data



Popularity

ranking



Biologists

are not
librarians


Research

Project
Hypothesis

http://www.hsls.pitt.edu/guides/genetics


OBRC
is

better

than

Google
at



Locating

bioinformatics

tools


Quickly



Easily



Successfully



Leaving

users

happier

with
:


Results


Overall

experience



Research

Project



http://www.hsls.pitt.edu/guides/genetics


10
biology

graduate

students



Three

tasks

on
both

OBRC and Google



Recorded

search

as
they

narrated




Search experience survey

Research

Project


Measured
:



Time
spent

on
each

search


Number

of
search

queries


Number

of sites
visited


Success

rate


Users
’ ratings


http://www.hsls.pitt.edu/guides/genetics

Task 1: Protein Interaction

http://www.hsls.pitt.edu/guides/genetics


You are a pharmacologist interested in
drugs that target proteins. Where would
you find reliable information about how
drugs interact with proteins, specifically
information about the strength of the
binding between the protein and
ligand
?


Task 2: Structure Prediction

http://www.hsls.pitt.edu/guides/genetics


Find a tool that performs protein structure
prediction. Determine from the search
engine how often it has been cited.


Task 3: Ovarian Cancer

http://www.hsls.pitt.edu/guides/genetics


Where would you look for information
about all the published gene mutations
implicated in ovarian cancer, in one place?


Is it faster?

http://www.hsls.pitt.edu/guides/genetics

0
50
100
150
200
250
300
350
Task 1
Task 2
Task 3
Average
Time (s)

Average Task Completion Time

Google
OBRC
Protein

Interaction

Structure

Prediction

Ovarian

Cancer

Is it easier to find results?

http://www.hsls.pitt.edu/guides/genetics

0
1
2
3
4
Task 1
Task 2
Task 3
Average
Number of Queries

Average Number of Search Queries

Google
OBRC
Protein

Interaction

Structure

Prediction

Ovarian

Cancer

Is it easier to find results?

http://www.hsls.pitt.edu/guides/genetics

0
1
2
3
4
Task 1
Task 2
Task 3
Average
Number of Sites Visited

Number of Sites Visited Before Success

Google
OBRC
Protein

Interaction

Structure

Prediction

Ovarian

Cancer

Is it successful?

http://www.hsls.pitt.edu/guides/genetics

0%
20%
40%
60%
80%
100%
Task 1
Task 2
Task 3
Average
Percentage of Successful Searches

Rate of Success

Google
OBRC
Protein

Interaction

Structure

Prediction

Ovarian

Cancer

Do users like it?

http://www.hsls.pitt.edu/guides/genetics

1
2
3
4
5
6
7
Appropriateness of Results
Ease of Use
Overall Experience
Rating

Survey Results

Google
OBRC
What

do
they

say

about Google?

http://www.hsls.pitt.edu/guides/genetics

“I was just frustrated


[by Google].”

“Google was difficult


to navigate with all the


information on cancer.”

“[Google is] time

consuming unless

we know what we

are searching for.”

“I realized that


Google is not an

effective search tool.”

What do they say about OBRC?

http://www.hsls.pitt.edu/guides/genetics

“Wow, this is

really direct!”

“Gives good results,

but I don’t like

the interface.”

“OBRC is an excellent

resource, especially if

you don’t know

exactly what you’re

searching for.”

“[OBRC] was much

easier to locate

relevant information.”

What

did

we

learn

about OBRC?


Time
lag



Scopus

and Google
Scholar



Spell

check



Long
search

queries



Unaware

it

exists

http://www.hsls.pitt.edu/guides/genetics

Limitations


Small
sample

size



Potential

bias

from

tasks



Previous

experience

with

Google


http://www.hsls.pitt.edu/guides/genetics

Future
Work


Larger

sample

size



Other

user groups



User
-
chosen

task


http://www.hsls.pitt.edu/guides/genetics

Conclusions…

http://www.hsls.pitt.edu/guides/genetics


Searches using OBRC are:


Faster


Easier


More successful



Users are more satisfied.



Users plan to use OBRC in the future.


Conclusions…

http://www.hsls.pitt.edu/guides/genetics

…libraries can and
should compete!


http://www.hsls.pitt.edu/guides/genetics


This study is IRB approved.


Funding provided by: NLM Grant 3 T15
LM007059
-
23S1


Thanks to my collaborators, Dr. Carrie
Iwema

and Dr.
Ansuman

Chattopadhyay


Special thanks to Nancy Tannery and Dr.
Rebecca Crowley


Thank you!


Any questions?


Katrina Kurtz

Carrie
Iwema



Ansuman

Chattopadhyay

kak59@pitt.edu

iwema@pitt.edu

ansuman@pitt.edu


412
-
624
-
3508

412
-
648
-
6887 412
-
648
-
1297


http://www.hsls.pitt.edu/guides/genetics