A probabilistic framework to predict protein function from interaction data integrated with semantic knowledge

BMC Bioinformatics

Table 3 Localization prediction results for unknown proteins

unknown	predicted subcellular localization		confidence
	ID in MIPS	description
YER070W	755	mitochondria	90.7
YJR033C	750	nucleus	215.9
YJR091C	722	integral membrane/endomembranes	49.6
YJR091C	725	cytoplasm	213.2
YJR091C	770	vacuole	52.1
YLL038C	705	bud	50.0
YML023C	722	integral membrane/endomembranes	119.8
YML023C	750	nucleus	191.3
YNL293W	705	bud	81.0
YNL293W	715	cell periphery	69.6
YNL293W	730	cytoskeleton	54.3
YOR076C	750.05	nucleolus	215.8
YOR076C	755	mitochondria	60.3
YOR231W	705	bud	168.7
YOR231W	715	cell periphery	58.0
YOR231W	730	cytoskeleton	45.3
YOR231W	750	nucleus	88.0

For each unknown protein, our probabilistic approach predicted a set of subcellular components with the prediction confidence, log λ. The prediction results, whose confidence is greater than 40, were listed. A protein can have more than one subcellular component predicted with different confidence rates.

ISSN: 1471-2105