Welcome to the OmniSearch Project
Funded by National Institutes of Health (NIH)/NCI
microRNAs (short for miRNAs or miRs) are a special class of small, non-coding RNA
molecules and have been reported to perform important roles in various biological
and pathological processes by regulating their respective target genes (short for
targets). As such, miRNAs are closely associated with the development, diagnosis,
and prognosis of various human diseases including cancer. In fact, prior research
has demonstrated that miRNAs may provide critical insights with regard to many aspects
of human diseases, including early diagnosis, personalized treatment, prognosis
prediction, and so forth.
However, miRNA knowledge acquisition still remains challenging despite of many research
efforts in this area. To completely understand and fully delineate miRNA functions,
besides performing direct biological experiments in "wet" labs, biologists and bioinformaticians
can also query PubMed and TarBase for biologically validated miRNA targets, and
various miRNA target prediction databases or websites for computationally predicted
targets as well. There exist two significant barriers in this scenario: (1) The
number of distinct miRNA target prediction databases is in the neighborhood of 30.
Moreover, different databases utilize different prediction algorithms, and more
importantly, these databases have, more often than not, quite heterogeneous semantics
(that is, the meaning of data) among each other. (2) Each individual miRNA may target
on up to hundreds or even more genes. For each and every target gene, either biologically
validated or computationally putative, it is often necessary to extract and obtain
additional information from other data sources, such as gene expression, protein
functions, and affiliated signaling pathways, because such additional information
is critical in helping biologists to better explore functions performed by miRNAs.
Similarly, these involved data sources are inherently heterogeneous with each other.
In short, biologists and bioinformaticians are facing significant barriers in fully
delineating miRNA functions and the following effective bio-curation.
To tackle the above-mentioned challenge, we will develop OmniSearch, a semantic
search tool, to assist biologists, cancer biologists in particular, in unraveling
critical roles of miRNAs in human cancers in an automated and highly efficient manner.
We will handle the significant challenge of data sharing, data integration, and
effective search in miRNA research in human oncology.