FunDO Frequently Asked Questions
How come I can't see the results?
We are using the Google Visualization API to display the results table. If you cannot see the table, try using another browser. Firefox, Safari, and Chrome are all known to render the table properly. Any other standards compliant browser is also expected to work.
Where can I download the data?
See here.
How does FunDO compare to DAVID/EASE?
FunDO provides analysis of disease terms associated with genes in a gene list, similar to Gene Ontology analysis via DAVID/EASE.
 
statistical test
annotation database
goal
developer
FunDO
hypergeometric test
Disease Ontology
translate findings into clinical care
NUBIC, Northwestern University
DAVID/EASE
hypergeometric test
Gene Ontology
interpret biological function
NIAID, NIH
Where do gene lists come from?
FunDO uses gene lists identified from complex data sets to explore the disease terms associated with the genes of interest.
How can I convert gene identifiers to the proper format?
FunDO accepts Entrez gene IDs (for human genes) or gene symbols. Gene identifiers should be whitespace or comma separated. Many statistical analysis packages for microarray data can directly output this format.

If you have identifiers in another format, we suggest using the Gene ID Conversion Tool from DAVID.
How do the statistics work?
We use a hypergeometic test to assess the p-value of the enrichment. In our case, this is equivalent to Fisher's exact test (one-sided).
What is the difference between GO analysis and DO analysis?
Here we use a gene list from a pancreatic cancer study (Iacobuzio-Donahue et al., American Journal of Pathology 2003; 162:1151-1162) to illustrate the difference between GO and DO analysis. Briefly, a list of 125 genes identified in that study was used for further functional analysis. The GO analysis was conducted using DAVID whereas the DO analysis was conducted using FunDO.

As expected, results from GO analysis suggest fundamental biological processes, such as protein binding, structural molecular activity, and structural constituents of cytoskeleton. In contrast, results from GO analysis suggest disease associations, such as cancer metastasis, sarcoma, breast, lung, colon, and pancreatic caner.

In particular, the FunDO analysis suggests the significance of 23 genes associated with cancer metastasis, which was not discussed in the original paper. This new finding provides clues to potentially new molecular markers for the surveillance and treatment of pancreatic cancer.
Gene Ontology Analysis using DAVID
protein binding
structural molecule activity
structural constituent of cytoskeleton
cytoplasm
pigment granule
melanosome
intracellular non-membrane-bound organelle
non-membrane-bound organelle
cytoskeleton
identical protein binding
 
Disease Ontology Analysis using FunDO
Neoplasm Metastasis
Sarcoma
Breast cancer
Lung cancer
Colon cancer
Stomach cancer
Pancreatic cancer
Squamous cell carcinoma
Prostate cancer
Ovarian cancer
Count
69
16
6
55
6
6
25
25
17
9
 
Count
23
25
26
20
18
17
14
13
17
13
%
69.00
16.00
6.00
55.00
6.00
6.00
25.00
25.00
17.00
9.00
 
%
22.77
24.75
25.74
19.8
17.82
16.83
13.86
12.87
16.83
12.87
Enrichment
1.73
3.23
12.45
1.51
12.18
12.18
2.18
2.18
2.54
4.26
 
Enrichment
22.88
16.56
14.32
20.30
22.37
21.46
21.48
22.25
11.65
18.72
p-value
1.96E-09
9.98E-05
1.15E-04
1.18E-04
1.27E-04
1.27E-04
2.45E-04
2.45E-04
8.08E-04
1.16E-03
 
p-value
7.65E-25
1.56E-23
7.01E-23
1.24E-20
2.18E-19
4.72E-18
4.92E-15
3.13E-14
1.06E-13
2.85E-13