Functional Annotation Routines Used by ABRF Bioinformatics Core Facilities - Observations, Comparisons, and...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Functional annotation of gene lists is crucial for genomics experiments.
• Bioinformatics core facilities require standardized approaches for this analysis.
• A lack of consensus exists in preferred functional annotation methods.

Purpose of the Study:

• To investigate and compare the performance of different functional annotation tools.
• To identify consensus and discrepancies in functional annotation approaches.
• To provide guidance for bioinformatics core facilities.

Main Methods:

• Selected 4 experiments with diverse designs.
• Analyzed gene sets from these experiments using 6 common bioinformatics tools.
• Focused on mapping results to annotation categories provided by each tool.
• Investigated Fisher's exact test parameters for optimization.

Main Results:

• Most tools successfully identified the selected biological results.
• Parameter adjustments, particularly for Fisher's exact test, resolved exceptions.
• Background set size had minimal impact; gene counts within categories and total genes were critical.
• Differences in annotation category composition and testing significantly affected results.

Conclusions:

• Functional annotation tools are generally effective but require careful parameter tuning.
• Understanding tool-specific annotation categories is essential for accurate interpretation.
• Standardization of functional annotation methods remains an area for development in bioinformatics cores.