UPEFinder

UPEFinder infers functional annotations for uncharacterized human proteins (uPE1 proteins and uncharacterized missing proteins, uMPs) by analyzing public RNA-Seq coexpression data and network topology using a guilt-by-association approach and PageRank.

Key Features:

• Guilt-by-Association Analysis: Leverages public RNA-Seq datasets to correlate uPE1 proteins and uMPs with PE1 proteins and construct a protein coexpression network.
• PageRank Algorithm Application: Applies the PageRank algorithm to the constructed network to prioritize nodes indicative of potential biological annotations based on gene expression correlations.
• Database Storage: Stores inferred annotations and network data in a dedicated database.

Scientific Applications:

• Functional Annotation of Uncharacterized Proteins: Provides inferred biological annotations for uPE1 proteins and uMPs to support characterization efforts.
• Hypothesis Generation for Protein Function and Disease Implications: Generates hypotheses about protein functions and their potential implications in human diseases using gene expression and network analysis.
• Support for the Human Proteome Project (HPP): Assists HPP efforts to characterize missing and uncharacterized proteins through systematic annotation from expression-derived networks.

Methodology:

Uses public RNA-Seq datasets to compute correlations between uncharacterized proteins and PE1 proteins, constructs a protein coexpression network, applies PageRank to prioritize nodes, and stores results in a dedicated database.