Transcript Presentation

Making “Open Data” Work:
Challenges for Data
Integration in Genomics
Research
Irene Pasquetto
@UCLA_KI
@irenepasquetto
1
Literature on OD in SCIENCE
2
THE CRANIOFACIAL
RESEARCH FIELD
• Interdisciplinary domain at the intersection
of biomedicine and pure biology research.
GOALS:
• Study the genetic causes of facial variation
and facial abnormalities.
• Study the evolutionary processes involved
in craniofacial development.
• Develop awareness, prevention and
treatments for common genetic syndromes
involving the face, such as cleft palate (half
of birth defects involves the face)
The Wonders of the East, Beowulf Manuscript, c. 700–1000 AD
3
4
DATA INTEGRATION IS
NECESSARY TO ALLOW
ANALYSIS AND REUSE, BUT
DIFFICULT BECAUSE:
• Data are collected from 4
different animal models (chimps,
mice, zebrafish and humans)
• Variety of data formats: 3D
images, gene expression data,
chip-seq, RNA-seq etc.
• Data collected and analyzed with
different methods (from single
genes experiments, to whole
genomics approaches)
LAB
10
LAB
1
LAB
2
LAB
9
LAB
3
INFORMATICS
HUB
LAB
8
LAB
4
LAB
7
LAB
6
LAB
5
5
What does “data integration” mean?
6
Conclusions
• Data reuse depends on the possibility of
conducting integrated data analysis.
• Data integration work is complicated by the high
heterogeneity of the datasets, methods, and tools.
• Negotiation of the meaning of “data integration”
(not just about standards!)
• Data integration work is emergent and vital for data
reuse, but it is difficult to articulate.
7
Thank you!
@irenepasquetto
@UCLA_KI
KI website: https://knowledgeinfrastructures.gseis.ucla.edu/
8