INVESTIGADORES
LLERA Andrea Sabina
congresos y reuniones científicas
Título:
Is there strength in numbers? Looking for the best background reference in ontology
Autor/es:
CRISTOBAL FRESNO; ANDREA LLERA; MARIA ROMINA GIROTTI; MARIA PIA VALACCO; JUAN ANTONIO LÓPEZ; OSVALDO PODHAJCER; MONICA BALZARINI; FEDERICO PRADA; ELMER FERNÁNDEZ
Lugar:
Hinxton, Cambridge
Reunión:
Conferencia; Wellcome Trust Conference on Functional Genomics and Systems Biology 2011; 2011
Institución organizadora:
Wellcome Trust
Resumen:
Set enrichment analysis (SEA) is used to
identify enriched biological categories/terms within a high-throughput differential
expression experiments. This is performed by evaluating the proportion of
differentially expressed genes against a background reference (BR). However,
the choice of the appropriate BR is a perplexing problem and results may
depend on it (particularly for proteomics studies). Hence, in order to do SEA the
user has to choose a reference, usually the whole genome or the chip-gene list.
Nevertheless, potentially relevant terms could be missing due to the BR choice.
In this work, a visualization procedure that helps SEA interpretation is
presented. The multi-reference contrast method (MRCM) combines simultaneous
results from multiple BRs (the genome, the chip-gene list and a reference lead
by the experimental setting). The MRCM facilitates the experimental
interpretation and also helps to validate relevant terms among consensus and/or
to identify potentially relevant terms by contrast.
The MRCM facilitates the exploration task involved
in ontology analysis on proteomic/genomic experiments, where consensus terms
were found to validate main experimental hypothesis. It also allows finding new
biologically relevant categories/terms (literature validated) in several various
experiments. In this sense, the use of more than one reference may provide new biological
insights. Non-consensus terms were automatically highlighted by the MRCM, helping
its interpretation and analysis. Its utility is shown in one proteomic study
and evaluated in three microarrays based experiments.