prioritization according to phenotype description

Question

[Emma]

Hi all,

I am looking for prioritization tools which use the phenotype description as input. Most prioritization tools ask seed genes as input. But the reality is that we sometimes can not give out an accurate diagnosis. There are so many syndromes that have similar phenotypes. Therefore, it is hard to decide what genes can used as seed genes. Besides, prioritization tools seems to exploit the guilt-by-association principle in the protein interaction net-work to uncover novel disease-gene associations. If we got the diagnosis wrong from the very beginning, it means that we might never get the underlying genes.

I was wondering if there any tools that can use several key words (I mean the phenotype descriptions) and then automatically come up with candidate seed genes? And then use these seed genes to prioritize the input candidate genes?

Best,
Emma

Answer 1

[Jean-Karim Heriche]

Gene prioritization tools (see a list here) don't just exploit protein-protein interactions but also other data types depending on the tool although, for human, protein-protein interactions usually constitute the most informative data set. Also not all tools use the guilt-by-association principle but this is what seems to be the most powerful approach.
Concerning prioritization directly by phenotype, have a look at Phenolyzer: paper, website, code.